Security Guide - A Guide to Securing Fedora Linux

Fedora 13 Security Guide A Guide to Securing Fedora Linux

Johnray Fuller John Ha David O'Brien Scott Radvan Eric Christensen Adam Ligas

Security Guide

Fedora 13 Security Guide A Guide to Securing Fedora Linux Edition 13.0 Author Author Author Author Author Author

Johnray Fuller John Ha David O'Brien Scott Radvan Eric Christensen Adam Ligas

[email protected] [email protected] [email protected] [email protected] [email protected] [email protected]

Copyright © 2010 Red Hat, Inc. The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/. The original authors of this document, and Red Hat, designate the Fedora Project as the "Attribution Party" for purposes of CC-BY-SA. In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version. Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law. Red Hat, Red Hat Enterprise Linux, the Shadowman logo, JBoss, MetaMatrix, Fedora, the Infinity Logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries. For guidelines on the permitted uses of the Fedora trademarks, refer to https://fedoraproject.org/wiki/ Legal:Trademark_guidelines. Linux® is the registered trademark of Linus Torvalds in the United States and other countries. Java® is a registered trademark of Oracle and/or its affiliates. XFS® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries. All other trademarks are the property of their respective owners.

The Fedora Security Guide is designed to assist users of Fedora in learning the processes and practices of securing workstations and servers against local and remote intrusion, exploitation, and malicious activity. Focused on Fedora Linux but detailing concepts and techniques valid for all Linux systems, the Fedora Security Guide details the planning and the tools involved in creating a secured computing environment for the YPXFRD_ARGS="-p 835"

The following iptables rules can then be used to enforce which network the server listens to for these ports: iptables -A INPUT -p ALL -s! 192.168.0.0/24 iptables -A INPUT -p ALL -s! 192.168.0.0/24

--dport 834 -j DROP --dport 835 -j DROP

This means that the server only allows connections to ports 834 and 835 if the requests come from the 192.168.0.0/24 network, regardless of the protocol.

Note Refer to Section 2.8, “Firewalls” for more information about implementing firewalls with iptables commands.

2.2.3.5. Use Kerberos Authentication One of the issues to consider when NIS is used for authentication is that whenever a user logs into a machine, a password hash from the /etc/shadow map is sent over the network. If an intruder gains access to an NIS domain and sniffs network traffic, they can collect usernames and password hashes.

45

Chapter 2. Securing Your Network

With enough time, a password cracking program can guess weak passwords, and an attacker can gain access to a valid account on the network. Since Kerberos uses secret-key cryptography, no password hashes are ever sent over the network, making the system far more secure. Refer to Section 2.6, “Kerberos” for more information about Kerberos.

2.2.4. Securing NFS Important The version of NFS included in Fedora, NFSv4, no longer requires the portmap service as outlined in Section 2.2.2, “Securing Portmap”. NFS traffic now utilizes TCP in all versions, rather than UDP, and requires it when using NFSv4. NFSv4 now includes Kerberos user and group authentication, as part of the RPCSEC_GSS kernel module. Information on portmap is still included, since Fedora supports NFSv2 and NFSv3, both of which utilize portmap.

2.2.4.1. Carefully Plan the Network Now that NFSv4 has the ability to pass all information encrypted using Kerberos over a network, it is important that the service be configured correctly if it is behind a firewall or on a segmented network. NFSv2 and NFSv3 still pass data insecurely, and this should be taken into consideration. Careful network design in all of these regards can help prevent security breaches.

2.2.4.2. Beware of Syntax Errors The NFS server determines which file systems to export and which hosts to export these directories to by consulting the /etc/exports file. Be careful not to add extraneous spaces when editing this file. For instance, the following line in the /etc/exports file shares the directory /tmp/nfs/ to the host bob.example.com with read/write permissions. /tmp/nfs/

bob.example.com(rw)

The following line in the /etc/exports file, on the other hand, shares the same directory to the host bob.example.com with read-only permissions and shares it to the world with read/write permissions due to a single space character after the hostname. /tmp/nfs/

bob.example.com (rw)

It is good practice to check any configured NFS shares by using the showmount command to verify what is being shared: showmount -e

2.2.4.3. Do Not Use the no_root_squash Option By default, NFS shares change the root user to the nfsnobody user, an unprivileged user account. This changes the owner of all root-created files to nfsnobody, which prevents uploading of programs with the setuid bit set.

46

Securing the Apache HTTP Server

If no_root_squash is used, remote root users are able to change any file on the shared file system and leave applications infected by trojans for other users to inadvertently execute.

2.2.4.4. NFS Firewall Configuration The ports used for NFS are assigned dynamically by rpcbind, which can cause problems when creating firewall rules. To simplify this process, use the /etc/sysconfig/nfs file to specify which ports are to be used: • MOUNTD_PORT — TCP and UDP port for mountd (rpc.mountd) • STATD_PORT — TCP and UDP port for status (rpc.statd) • LOCKD_TCPPORT — TCP port for nlockmgr (rpc.lockd) • LOCKD_UDPPORT — UDP port nlockmgr (rpc.lockd) Port numbers specified must not be used by any other service. Configure your firewall to allow the port numbers specified, as well as TCP and UDP port 2049 (NFS). Run the rpcinfo -p command on the NFS server to see which ports and RPC programs are being used.

2.2.5. Securing the Apache HTTP Server The Apache HTTP Server is one of the most stable and secure services that ships with Fedora. A large number of options and techniques are available to secure the Apache HTTP Server — too numerous to delve into deeply here. The following section briefly explains good practices when running the Apache HTTP Server. Always verify that any scripts running on the system work as intended before putting them into production. Also, ensure that only the root user has write permissions to any directory containing scripts or CGIs. To do this, run the following commands as the root user: 1.

chown root

2.

chmod 755

System administrators should be careful when using the following configuration options (configured in /etc/httpd/conf/httpd.conf): FollowSymLinks This directive is enabled by default, so be sure to use caution when creating symbolic links to the document root of the Web server. For instance, it is a bad idea to provide a symbolic link to /. Indexes This directive is enabled by default, but may not be desirable. To prevent visitors from browsing files on the server, remove this directive. UserDir The UserDir directive is disabled by default because it can confirm the presence of a user account on the system. To enable user directory browsing on the server, use the following directives:

47


UserDir enabled UserDir disabled root

These directives activate user directory browsing for all user directories other than /root/. To add users to the list of disabled accounts, add a space-delimited list of users on the UserDir disabled line.

Important Do not remove the IncludesNoExec directive. By default, the Server-Side Includes (SSI) module cannot execute commands. It is recommended that you do not change this setting unless absolutely necessary, as it could, potentially, enable an attacker to execute commands on the system.

2.2.6. Securing FTP The File Transfer Protocol (FTP) is an older TCP protocol designed to transfer files over a network. Because all transactions with the server, including user authentication, are unencrypted, it is considered an insecure protocol and should be carefully configured. Fedora provides three FTP servers. • gssftpd — A Kerberos-aware xinetd-based FTP daemon that does not transmit authentication information over the network. • Red Hat Content Accelerator (tux) — A kernel-space Web server with FTP capabilities. • vsftpd — A standalone, security oriented implementation of the FTP service. The following security guidelines are for setting up the vsftpd FTP service.

2.2.6.1. FTP Greeting Banner Before submitting a username and password, all users are presented with a greeting banner. By default, this banner includes version information useful to crackers trying to identify weaknesses in a system. To change the greeting banner for vsftpd, add the following directive to the /etc/vsftpd/ vsftpd.conf file: ftpd_banner=

Replace in the above directive with the text of the greeting message. For mutli-line banners, it is best to use a banner file. To simplify management of multiple banners, place all banners in a new directory called /etc/banners/. The banner file for FTP connections in this example is /etc/banners/ftp.msg. Below is an example of what such a file may look like: ######### # Hello, all activity on ftp.example.com is logged. #########

48

Securing FTP

Note It is not necessary to begin each line of the file with 220 as specified in Section 2.2.1.1.1, “TCP Wrappers and Connection Banners”. To reference this greeting banner file for vsftpd, add the following directive to the /etc/vsftpd/ vsftpd.conf file: banner_file=/etc/banners/ftp.msg

It also is possible to send additional banners to incoming connections using TCP Wrappers as described in Section 2.2.1.1.1, “TCP Wrappers and Connection Banners”.

2.2.6.2. Anonymous Access The presence of the /var/ftp/ directory activates the anonymous account. The easiest way to create this directory is to install the vsftpd package. This package establishes a directory tree for anonymous users and configures the permissions on directories to read-only for anonymous users. By default the anonymous user cannot write to any directories.

Warning If enabling anonymous access to an FTP server, be aware of where sensitive data is stored.

2.2.6.2.1. Anonymous Upload To allow anonymous users to upload files, it is recommended that a write-only directory be created within /var/ftp/pub/. To do this, type the following command: mkdir /var/ftp/pub/upload

Next, change the permissions so that anonymous users cannot view the contents of the directory: chmod 730 /var/ftp/pub/upload

A long format listing of the directory should look like this: drwx-wx---

2 root

ftp

4096 Feb 13 20:05 upload

Warning Administrators who allow anonymous users to read and write in directories often find that their servers become a repository of stolen software.

49


Additionally, under vsftpd, add the following line to the /etc/vsftpd/vsftpd.conf file: anon_upload_enable=YES

2.2.6.3. User Accounts Because FTP transmits unencrypted usernames and passwords over insecure networks for authentication, it is a good idea to deny system users access to the server from their user accounts. To disable all user accounts in vsftpd, add the following directive to /etc/vsftpd/vsftpd.conf: local_enable=NO

2.2.6.3.1. Restricting User Accounts To disable FTP access for specific accounts or specific groups of accounts, such as the root user and those with sudo privileges, the easiest way is to use a PAM list file as described in Section 2.1.4.2.4, “Disabling Root Using PAM”. The PAM configuration file for vsftpd is /etc/pam.d/vsftpd. It is also possible to disable user accounts within each service directly. To disable specific user accounts in vsftpd, add the username to /etc/vsftpd.ftpusers

2.2.6.4. Use TCP Wrappers To Control Access Use TCP Wrappers to control access to either FTP daemon as outlined in Section 2.2.1.1, “Enhancing Security With TCP Wrappers”.

2.2.7. Securing Sendmail Sendmail is a Mail Transfer Agent (MTA) that uses the Simple Mail Transfer Protocol (SMTP) to deliver electronic messages between other MTAs and to email clients or delivery agents. Although many MTAs are capable of encrypting traffic between one another, most do not, so sending email over any public networks is considered an inherently insecure form of communication. It is recommended that anyone planning to implement a Sendmail server address the following issues.

2.2.7.1. Limiting a Denial of Service Attack Because of the nature of email, a determined attacker can flood the server with mail fairly easily and cause a denial of service. By setting limits to the following directives in /etc/mail/sendmail.mc, the effectiveness of such attacks is limited. • confCONNECTION_RATE_THROTTLE — The number of connections the server can receive per second. By default, Sendmail does not limit the number of connections. If a limit is set and reached, further connections are delayed. • confMAX_DAEMON_CHILDREN — The maximum number of child processes that can be spawned by the server. By default, Sendmail does not assign a limit to the number of child processes. If a limit is set and reached, further connections are delayed. • confMIN_FREE_BLOCKS — The minimum number of free blocks which must be available for the server to accept mail. The default is 100 blocks.

50

Verifying Which Ports Are Listening

• confMAX_HEADERS_LENGTH — The maximum acceptable size (in bytes) for a message header. • confMAX_MESSAGE_SIZE — The maximum acceptable size (in bytes) for a single message.

2.2.7.2. NFS and Sendmail Never put the mail spool directory, /var/spool/mail/, on an NFS shared volume. Because NFSv2 and NFSv3 do not maintain control over user and group IDs, two or more users can have the same UID, and receive and read each other's mail.

Note With NFSv4 using Kerberos, this is not the case, since the SECRPC_GSS kernel module does not utilize UID-based authentication. However, it is still considered good practice not to put the mail spool directory on NFS shared volumes.

2.2.7.3. Mail-only Users To help prevent local user exploits on the Sendmail server, it is best for mail users to only access the Sendmail server using an email program. Shell accounts on the mail server should not be allowed and all user shells in the /etc/passwd file should be set to /sbin/nologin (with the possible exception of the root user).

2.2.8. Verifying Which Ports Are Listening After configuring network services, it is important to pay attention to which ports are actually listening on the system's network interfaces. Any open ports can be evidence of an intrusion. There are two basic approaches for listing the ports that are listening on the network. The less reliable approach is to query the network stack using commands such as netstat -an or lsof -i. This method is less reliable since these programs do not connect to the machine from the network, but rather check to see what is running on the system. For this reason, these applications are frequent targets for replacement by attackers. Crackers attempt to cover their tracks if they open unauthorized network ports by replacing netstat and lsof with their own, modified versions. A more reliable way to check which ports are listening on the network is to use a port scanner such as nmap. The following command issued from the console determines which ports are listening for TCP connections from the network: nmap -sT -O localhost

The output of this command appears as follows: Starting Nmap 4.68 ( http://nmap.org ) at 2009-03-06 12:08 EST Interesting ports on localhost.localdomain (127.0.0.1): Not shown: 1711 closed ports PORT STATE SERVICE 22/tcp open ssh 25/tcp open smtp

51


111/tcp open rpcbind 113/tcp open auth 631/tcp open ipp 834/tcp open unknown 2601/tcp open zebra 32774/tcp open sometimes-rpc11 Device type: general purpose Running: Linux 2.6.X OS details: Linux 2.6.17 - 2.6.24 Uptime: 4.122 days (since Mon Mar 2 09:12:31 2009) Network Distance: 0 hops OS detection performed. Please report any incorrect results at http://nmap.org/submit/ . Nmap done: 1 IP address (1 host up) scanned in 1.420 seconds

This output shows the system is running portmap due to the presence of the sunrpc service. However, there is also a mystery service on port 834. To check if the port is associated with the official list of known services, type: cat /etc/services | grep 834

This command returns no output. This indicates that while the port is in the reserved range (meaning 0 through 1023) and requires root access to open, it is not associated with a known service. Next, check for information about the port using netstat or lsof. To check for port 834 using netstat, use the following command: netstat -anp | grep 834

The command returns the following output: tcp

0

0 0.0.0.0:834

0.0.0.0:*

LISTEN

653/ypbind

The presence of the open port in netstat is reassuring because a cracker opening a port surreptitiously on a hacked system is not likely to allow it to be revealed through this command. Also, the [p] option reveals the process ID (PID) of the service that opened the port. In this case, the open port belongs to ypbind (NIS), which is an RPC service handled in conjunction with the portmap service. The lsof command reveals similar information to netstat since it is also capable of linking open ports to services: lsof -i | grep 834

The relevant portion of the output from this command follows: ypbind ypbind ypbind ypbind

653 655 656 657

0 0 0 0

7u 7u 7u 7u

IPv4 IPv4 IPv4 IPv4

1319 1319 1319 1319

TCP TCP TCP TCP

*:834 *:834 *:834 *:834

(LISTEN) (LISTEN) (LISTEN) (LISTEN)

These tools reveal a great deal about the status of the services running on a machine. These tools are flexible and can provide a wealth of information about network services and configuration. Refer to the man pages for lsof, netstat, nmap, and services for more information.

52

Single Sign-on (SSO)

2.3. Single Sign-on (SSO) 2.3.1. Introduction The Fedora SSO functionality reduces the number of times Fedora desktop users have to enter their passwords. Several major applications leverage the same underlying authentication and authorization mechanisms so that users can log in to Fedora from the log-in screen, and then not need to re-enter their passwords. These applications are detailed below. In addition, users can log in to their machines even when there is no network (offline mode) or where network connectivity is unreliable, for example, wireless access. In the latter case, services will degrade gracefully.

2.3.1.1. Supported Applications The following applications are currently supported by the unified log-in scheme in Fedora: • Login • Screensaver • Firefox and Thunderbird

2.3.1.2. Supported Authentication Mechanisms Fedora currently supports the following authentication mechanisms: • Kerberos name/password login • Smart card/PIN login

2.3.1.3. Supported Smart Cards Fedora has been tested with the Cyberflex e-gate card and reader, but any card that complies with both Java card 2.1.1 and Global Platform 2.0.1 specifications should operate correctly, as should any reader that is supported by PCSC-lite. Fedora has also been tested with Common Access Cards (CAC). The supported reader for CAC is the SCM SCR 331 USB Reader. As of Fedora 5.2, Gemalto smart cards (Cyberflex Access 64k v2, standard with DER SHA1 value configured as in PKCSI v2.1) are now supported. These smart cards now use readers compliant with Chip/Smart Card Interface Devices (CCID).

2.3.1.4. Advantages of Fedora Single Sign-on Numerous security mechanisms currently exist that utilize a large number of protocols and credential stores. Examples include SSL, SSH, IPsec, and Kerberos. Fedora SSO aims to unify these schemes to support the requirements listed above. This does not mean replacing Kerberos with X.509v3 certificates, but rather uniting them to reduce the burden on both system users and the administrators who manage them. To achieve this goal, Fedora:

53


• Provides a single, shared instance of the NSS crypto libraries on each operating system. • Ships the Certificate System's Enterprise Security Client (ESC) with the base operating system. The ESC application monitors smart card insertion events. If it detects that the user has inserted a smart card that was designed to be used with the Fedora Certificate System server product, it displays a user interface instructing the user how to enroll that smart card. • Unifies Kerberos and NSS so that users who log in to the operating system using a smart card also obtain a Kerberos credential (which allows them to log in to file servers, etc.)

2.3.2. Getting Started with your new Smart Card Before you can use your smart card to log in to your system and take advantage of the increased security options this technology provides, you need to perform some basic installation and configuration steps. These are described below.

Note This section provides a high-level view of getting started with your smart card. More detailed information is available in the Red Hat Certificate System Enterprise Security Client Guide. 1.

Log in with your Kerberos name and password

2.

Make sure you have the nss-tools package loaded.

3.

Download and install your corporate-specific root certificates. Use the following command to install the root CA certificate: certutil -A -d /etc/pki/nssdb -n "root ca cert" -t "CT,C,C" -i ./ ca_cert_in_base64_format.crt

4.

Verify that you have the following RPMs installed on your system: esc, pam_pkcs11, coolkey, ifdegate, ccid, gdm, authconfig, and authconfig-gtk.

5.

Enable Smart Card Login Support a.

On the Gnome Title Bar, select System->Administration->Authentication.

b.

Type your machine's root password if necessary.

c.

In the Authentication Configuration dialog, click the Authentication tab.

d.

Select the Enable Smart Card Support check box.

e.

Click the Configure Smart Card... button to display the Smartcard Settings dialog, and specify the required settings: • Require smart card for login — Clear this check box. After you have successfully logged in with the smart card you can select this option to prevent users from logging in without a smart card. • Card Removal Action — This controls what happens when you remove the smart card after you have logged in. The available options are:

54

How Smart Card Enrollment Works

• Lock — Removing the smart card locks the X screen. • Ignore — Removing the smart card has no effect. 6.

If you need to enable the Online Certificate Status Protocol (OCSP), open the /etc/ pam_pkcs11/pam_pkcs11.conf file, and locate the following line: enable_ocsp = false; Change this value to true, as follows: enable_ocsp = true;

7.

Enroll your smart card

8.

If you are using a CAC card, you also need to perform the following steps: a.

Change to the root account and create a file called /etc/pam_pkcs11/cn_map.

b.

Add the following entry to the cn_map file: MY.CAC_CN.123454 -> myloginid where MY.CAC_CN.123454 is the Common Name on your CAC and myloginid is your UNIX login ID.

9.

Logout

2.3.2.1. Troubleshooting If you have trouble getting your smart card to work, try using the following command to locate the source of the problem: pklogin_finder debug

If you run the pklogin_finder tool in debug mode while an enrolled smart card is plugged in, it attempts to output information about the validity of certificates, and if it is successful in attempting to map a login ID from the certificates that are on the card.

2.3.3. How Smart Card Enrollment Works Smart cards are said to be enrolled when they have received an appropriate certificate signed by a valid Certificate Authority (CA). This involves several steps, described below: 1. The user inserts their smart card into the smart card reader on their workstation. This event is recognized by the Enterprise Security Client (ESC). 2. The enrollment page is displayed on the user's desktop. The user completes the required details and the user's system then connects to the Token Processing System (TPS) and the CA. 3. The TPS enrolls the smart card using a certificate signed by the CA.

55


Figure 2.4. How Smart Card Enrollment Works

2.3.4. How Smart Card Login Works This section provides a brief overview of the process of logging in using a smart card. 1. When the user inserts their smart card into the smart card reader, this event is recognized by the PAM facility, which prompts for the user's PIN. 2. The system then looks up the user's current certificates and verifies their validity. The certificate is then mapped to the user's UID. 3. This is validated against the KDC and login granted.

56

Configuring Firefox to use Kerberos for SSO

Figure 2.5. How Smart Card Login Works

Note You cannot log in with a card that has not been enrolled, even if it has been formatted. You need to log in with a formatted, enrolled card, or not using a smart card, before you can enroll a new card. Refer to Section 2.6, “Kerberos” and Section 2.4, “Pluggable Authentication Modules (PAM)” for more information on Kerberos and PAM.

2.3.5. Configuring Firefox to use Kerberos for SSO You can configure Firefox to use Kerberos for Single Sign-on. In order for this functionality to work correctly, you need to configure your web browser to send your Kerberos credentials to the appropriate KDC.The following section describes the configuration changes and other requirements to achieve this. 1. In the address bar of Firefox, type about:config to display the list of current configuration options. 2. In the Filter field, type negotiate to restrict the list of options. 3. Double-click the network.negotiate-auth.trusted-uris entry to display the Enter string value dialog box.

57


4. Enter the name of the domain against which you want to authenticate, for example, .example.com. 5. Repeat the above procedure for the network.negotiate-auth.delegation-uris entry, using the same domain.

Note You can leave this value blank, as it allows Kerberos ticket passing, which is not required. If you do not see these two configuration options listed, your version of Firefox may be too old to support Negotiate authentication, and you should consider upgrading.

Figure 2.6. Configuring Firefox for SSO with Kerberos You now need to ensure that you have Kerberos tickets. In a command shell, type kinit to retrieve Kerberos tickets. To display the list of available tickets, type klist. The following shows an example output from these commands: [user@host ~] $ kinit Password for [email protected]: [user@host ~] $ klist Ticket cache: FILE:/tmp/krb5cc_10920 Default principal: [email protected] Valid starting Expires Service principal 10/26/06 23:47:54 10/27/06 09:47:54 krbtgt/[email protected] renew until 10/26/06 23:47:54 Kerberos 4 ticket cache: /tmp/tkt10920 klist: You have no tickets cached

58

Pluggable Authentication Modules (PAM)

2.3.5.1. Troubleshooting If you have followed the configuration steps above and Negotiate authentication is not working, you can turn on verbose logging of the authentication process. This could help you find the cause of the problem. To enable verbose logging, use the following procedure: 1. Close all instances of Firefox. 2. Open a command shell, and enter the following commands: export NSPR_LOG_MODULES=negotiateauth:5 export NSPR_LOG_FILE=/tmp/moz.log

3. Restart Firefox from that shell, and visit the website you were unable to authenticate to earlier. Information will be logged to /tmp/moz.log, and may give a clue to the problem. For example: -1208550944[90039d0]: entering nsNegotiateAuth::GetNextToken() -1208550944[90039d0]: gss_init_sec_context() failed: Miscellaneous failure No credentials cache found

This indicates that you do not have Kerberos tickets, and need to run kinit. If you are able to run kinit successfully from your machine but you are unable to authenticate, you might see something like this in the log file: -1208994096[8d683d8]: entering nsAuthGSSAPI::GetNextToken() -1208994096[8d683d8]: gss_init_sec_context() failed: Miscellaneous failure Server not found in Kerberos database

This generally indicates a Kerberos configuration problem. Make sure that you have the correct entries in the [domain_realm] section of the /etc/krb5.conf file. For example: .example.com = EXAMPLE.COM example.com = EXAMPLE.COM

If nothing appears in the log it is possible that you are behind a proxy, and that proxy is stripping off the HTTP headers required for Negotiate authentication. As a workaround, you can try to connect to the server using HTTPS instead, which allows the request to pass through unmodified. Then proceed to debug using the log file, as described above.

2.4. Pluggable Authentication Modules (PAM) Programs that grant users access to a system use authentication to verify each other's identity (that is, to establish that a user is who they say they are). Historically, each program had its own way of authenticating users. In Fedora, many programs are configured to use a centralized authentication mechanism called Pluggable Authentication Modules (PAM). PAM uses a pluggable, modular architecture, which affords the system administrator a great deal of flexibility in setting authentication policies for the system. In most situations, the default PAM configuration file for a PAM-aware application is sufficient. Sometimes, however, it is necessary to edit a PAM configuration file. Because misconfiguration of PAM can compromise system security, it is important to understand the structure of these files

59


before making any modifications. Refer to Section 2.4.3, “PAM Configuration File Format” for more information.

2.4.1. Advantages of PAM PAM offers the following advantages: • a common authentication scheme that can be used with a wide variety of applications. • significant flexibility and control over authentication for both system administrators and application developers. • a single, fully-documented library which allows developers to write programs without having to create their own authentication schemes.

2.4.2. PAM Configuration Files The /etc/pam.d/ directory contains the PAM configuration files for each PAM-aware application. In earlier versions of PAM, the /etc/pam.conf file was used, but this file is now deprecated and is only used if the /etc/pam.d/ directory does not exist.

2.4.2.1. PAM Service Files Each PAM-aware application or service has a file in the /etc/pam.d/ directory. Each file in this directory has the same name as the service to which it controls access. The PAM-aware program is responsible for defining its service name and installing its own PAM configuration file in the /etc/pam.d/ directory. For example, the login program defines its service name as login and installs the /etc/pam.d/login PAM configuration file.

2.4.3. PAM Configuration File Format Each PAM configuration file contains a group of directives formatted as follows:

Each of these elements is explained in the following sections.

2.4.3.1. Module Interface Four types of PAM module interface are currently available. Each of these corresponds to a different aspect of the authorization process: • auth — This module interface authenticates use. For example, it requests and verifies the validity of a password. Modules with this interface can also set credentials, such as group memberships or Kerberos tickets. • account — This module interface verifies that access is allowed. For example, it may check if a user account has expired or if a user is allowed to log in at a particular time of day. • password — This module interface is used for changing user passwords. • session — This module interface configures and manages user sessions. Modules with this interface can also perform additional tasks that are needed to allow access, like mounting a user's home directory and making the user's mailbox available.

60

PAM Configuration File Format

Note An individual module can provide any or all module interfaces. For instance, pam_unix.so provides all four module interfaces. In a PAM configuration file, the module interface is the first field defined. For example, a typical line in a configuration may look like this: auth required pam_unix.so

This instructs PAM to use the pam_unix.so module's auth interface.

2.4.3.1.1. Stacking Module Interfaces Module interface directives can be stacked, or placed upon one another, so that multiple modules are used together for one purpose. If a module's control flag uses the "sufficient" or "requisite" value (refer to Section 2.4.3.2, “Control Flag” for more information on these flags), then the order in which the modules are listed is important to the authentication process. Stacking makes it easy for an administrator to require specific conditions to exist before allowing the user to authenticate. For example, the reboot command normally uses several stacked modules, as seen in its PAM configuration file: [root@MyServer ~]# cat /etc/pam.d/reboot #%PAM-1.0 auth sufficient pam_rootok.so auth required pam_console.so #auth include system-auth account required pam_permit.so

• The first line is a comment and is not processed. • auth sufficient pam_rootok.so — This line uses the pam_rootok.so module to check whether the current user is root, by verifying that their UID is 0. If this test succeeds, no other modules are consulted and the command is executed. If this test fails, the next module is consulted. • auth required pam_console.so — This line uses the pam_console.so module to attempt to authenticate the user. If this user is already logged in at the console, pam_console.so checks whether there is a file in the /etc/security/console.apps/ directory with the same name as the service name (reboot). If such a file exists, authentication succeeds and control is passed to the next module. • #auth include system-auth — This line is commented and is not processed. • account required pam_permit.so — This line uses the pam_permit.so module to allow the root user or anyone logged in at the console to reboot the system.

2.4.3.2. Control Flag All PAM modules generate a success or failure result when called. Control flags tell PAM what do with the result. Modules can be stacked in a particular order, and the control flags determine how important the success or failure of a particular module is to the overall goal of authenticating the user to the service.

61


There are four predefined control flags: • required — The module result must be successful for authentication to continue. If the test fails at this point, the user is not notified until the results of all module tests that reference that interface are complete. • requisite — The module result must be successful for authentication to continue. However, if a test fails at this point, the user is notified immediately with a message reflecting the first failed required or requisite module test. • sufficient — The module result is ignored if it fails. However, if the result of a module flagged sufficient is successful and no previous modules flagged required have failed, then no other results are required and the user is authenticated to the service. • optional — The module result is ignored. A module flagged as optional only becomes necessary for successful authentication when no other modules reference the interface.

Important The order in which required modules are called is not critical. Only the sufficient and requisite control flags cause order to become important. A newer control flag syntax that allows for more precise control is now available for PAM. The pam.d man page, and the PAM documentation, located in the /usr/share/doc/ pam-/ directory, where is the version number for PAM on your system, describe this newer syntax in detail.

2.4.3.3. Module Name The module name provides PAM with the name of the pluggable module containing the specified module interface. In older versions of Fedora, the full path to the module was provided in the PAM configuration file. However, since the advent of multilib systems, which store 64-bit PAM modules in the /lib64/security/ directory, the directory name is omitted because the application is linked to the appropriate version of libpam, which can locate the correct version of the module.

2.4.3.4. Module Arguments PAM uses arguments to pass information to a pluggable module during authentication for some modules. For example, the pam_userdb.so module uses information stored in a Berkeley DB file to authenticate the user. Berkeley DB is an open source database system embedded in many applications. The module takes a db argument so that Berkeley DB knows which database to use for the requested service. The following is a typical pam_userdb.so line in a PAM configuration. The is the full path to the Berkeley DB database file: auth required pam_userdb.so db=

62

Sample PAM Configuration Files

Invalid arguments are generally ignored and do not otherwise affect the success or failure of the PAM module. Some modules, however, may fail on invalid arguments. Most modules report errors to the / var/log/secure file.

2.4.4. Sample PAM Configuration Files The following is a sample PAM application configuration file: #%PAM-1.0 auth required pam_securetty.so auth required pam_unix.so nullok auth required pam_nologin.so account required pam_unix.so password required pam_cracklib.so retry=3 password required pam_unix.so shadow nullok use_authtok session required pam_unix.so

• The first line is a comment, indicated by the hash mark (#) at the beginning of the line. • Lines two through four stack three modules for login authentication. auth required pam_securetty.so — This module ensures that if the user is trying to log in as root, the tty on which the user is logging in is listed in the /etc/securetty file, if that file exists. If the tty is not listed in the file, any attempt to log in as root fails with a Login incorrect message. auth required pam_unix.so nullok — This module prompts the user for a password and then checks the password using the information stored in /etc/passwd and, if it exists, /etc/ shadow. • The argument nullok instructs the pam_unix.so module to allow a blank password. • auth required pam_nologin.so — This is the final authentication step. It checks whether the /etc/nologin file exists. If it exists and the user is not root, authentication fails.

Note In this example, all three auth modules are checked, even if the first auth module fails. This prevents the user from knowing at what stage their authentication failed. Such knowledge in the hands of an attacker could allow them to more easily deduce how to crack the system. • account required pam_unix.so — This module performs any necessary account verification. For example, if shadow passwords have been enabled, the account interface of the pam_unix.so module checks to see if the account has expired or if the user has not changed the password within the allowed grace period. • password required pam_cracklib.so retry=3 — If a password has expired, the password component of the pam_cracklib.so module prompts for a new password. It then tests the newly created password to see whether it can easily be determined by a dictionary-based password cracking program.

63


• The argument retry=3 specifies that if the test fails the first time, the user has two more chances to create a strong password. • password required pam_unix.so shadow nullok use_authtok — This line specifies that if the program changes the user's password, it should use the password interface of the pam_unix.so module to do so. • The argument shadow instructs the module to create shadow passwords when updating a user's password. • The argument nullok instructs the module to allow the user to change their password from a blank password, otherwise a null password is treated as an account lock. • The final argument on this line, use_authtok, provides a good example of the importance of order when stacking PAM modules. This argument instructs the module not to prompt the user for a new password. Instead, it accepts any password that was recorded by a previous password module. In this way, all new passwords must pass the pam_cracklib.so test for secure passwords before being accepted. • session required pam_unix.so — The final line instructs the session interface of the pam_unix.so module to manage the session. This module logs the user name and the service type to /var/log/secure at the beginning and end of each session. This module can be supplemented by stacking it with other session modules for additional functionality.

2.4.5. Creating PAM Modules You can create or add new PAM modules at any time for use by PAM-aware applications. For example, a developer might create a one-time-password creation method and write a PAM module to support it. PAM-aware programs can immediately use the new module and password method without being recompiled or otherwise modified. This allows developers and system administrators to mix-and-match, as well as test, authentication methods for different programs without recompiling them. Documentation on writing modules is included in the /usr/share/doc/pam-/ directory, where is the version number for PAM on your system.

2.4.6. PAM and Administrative Credential Caching A number of graphical administrative tools in Fedora provide users with elevated privileges for up to five minutes using the pam_timestamp.so module. It is important to understand how this mechanism works, because a user who walks away from a terminal while pam_timestamp.so is in effect leaves the machine open to manipulation by anyone with physical access to the console. In the PAM timestamp scheme, the graphical administrative application prompts the user for the root password when it is launched. When the user has been authenticated, the pam_timestamp.so module creates a timestamp file. By default, this is created in the /var/run/sudo/ directory. If the timestamp file already exists, graphical administrative programs do not prompt for a password. Instead, the pam_timestamp.so module freshens the timestamp file, reserving an extra five minutes of unchallenged administrative access for the user.

64

PAM and Administrative Credential Caching

You can verify the actual state of the timestamp file by inspecting the /var/run/sudo/ file. For the desktop, the relevant file is unknown:root. If it is present and its timestamp is less than five minutes old, the credentials are valid. The existence of the timestamp file is indicated by an authentication icon, which appears in the notification area of the panel.

Figure 2.7. The Authentication Icon

2.4.6.1. Removing the Timestamp File Before abandoning a console where a PAM timestamp is active, it is recommended that the timestamp file be destroyed. To do this from a graphical environment, click the authentication icon on the panel. This causes a dialog box to appear. Click the Forget Authorization button to destroy the active timestamp file.

Figure 2.8. Dismiss Authentication Dialog You should be aware of the following with respect to the PAM timestamp file: • If logged in to the system remotely using ssh, use the /sbin/pam_timestamp_check -k root command to destroy the timestamp file. • You need to run the /sbin/pam_timestamp_check -k root command from the same terminal window from which you launched the privileged application. • You must be logged in as the user who originally invoked the pam_timestamp.so module in order to use the /sbin/pam_timestamp_check -k command. Do not log in as root to use this command. • If you want to kill the credentials on the desktop (without using the Forget Authorization action on the icon), use the following command: /sbin/pam_timestamp_check -k root /dev/null 2>/dev/null

Failure to use this command will only remove the credentials (if any) from the pty where you run the command. Refer to the pam_timestamp_check man page for more information about destroying the timestamp file using pam_timestamp_check.

65


2.4.6.2. Common pam_timestamp Directives The pam_timestamp.so module accepts several directives. The following are the two most commonly used options: • timestamp_timeout — Specifies the period (in seconds) for which the timestamp file is valid. The default value is 300 (five minutes). • timestampdir — Specifies the directory in which the timestamp file is stored. The default value is /var/run/sudo/. Refer to Section 2.8.9.1, “Installed Firewall Documentation” for more information about controlling the pam_timestamp.so module.

2.4.7. PAM and Device Ownership In Fedora, the first user who logs in at the physical console of the machine can manipulate certain devices and perform certain tasks normally reserved for the root user. This is controlled by a PAM module called pam_console.so.

2.4.7.1. Device Ownership When a user logs in to a Fedora system, the pam_console.so module is called by login or the graphical login programs, gdm, kdm, and xdm. If this user is the first user to log in at the physical console — referred to as the console user — the module grants the user ownership of a variety of devices normally owned by root. The console user owns these devices until the last local session for that user ends. After this user has logged out, ownership of the devices reverts back to the root user. The devices affected include, but are not limited to, sound cards, diskette drives, and CD-ROM drives. This facility allows a local user to manipulate these devices without obtaining root access, thus simplifying common tasks for the console user. You can modify the list of devices controlled by pam_console.so by editing the following files: • /etc/security/console.perms • /etc/security/console.perms.d/50-default.perms You can change the permissions of different devices than those listed in the above files, or override the specified defaults. Rather than modify the 50-default.perms file, you should create a new file (for example, xx-name.perms) and enter the required modifications. The name of the new default file must begin with a number higher than 50 (for example, 51-default.perms). This will override the defaults in the 50-default.perms file.

66

Additional Resources

Warning If the gdm, kdm, or xdm display manager configuration file has been altered to allow remote users to log in and the host is configured to run at runlevel 5, it is advisable to change the and directives in the /etc/security/ console.perms to the following values: =tty[0-9][0-9]* vc/[0-9][0-9]* :0\.[0-9] :0 =:0\.[0-9] :0

This prevents remote users from gaining access to devices and restricted applications on the machine. If the gdm, kdm, or xdm display manager configuration file has been altered to allow remote users to log in and the host is configured to run at any multiple user runlevel other than 5, it is advisable to remove the directive entirely and change the directive to the following value: =tty[0-9][0-9]* vc/[0-9][0-9]*

2.4.7.2. Application Access The console user also has access to certain programs configured for use in the /etc/security/ console.apps/ directory. This directory contains configuration files which enable the console user to run certain applications in /sbin and /usr/sbin. These configuration files have the same name as the applications that they set up. One notable group of applications that the console user has access to are three programs that shut down or reboot the system: • /sbin/halt • /sbin/reboot • /sbin/poweroff Because these are PAM-aware applications, they call the pam_console.so module as a requirement for use. Refer to Section 2.8.9.1, “Installed Firewall Documentation” for more information.

2.4.8. Additional Resources The following resources further explain methods to use and configure PAM. In addition to these resources, read the PAM configuration files on the system to better understand how they are structured.

67


2.4.8.1. Installed PAM Documentation • PAM-related man pages — Several man pages exist for the various applications and configuration files involved with PAM. The following is a list of some of the more important man pages. Configuration Files • pam — Good introductory information on PAM, including the structure and purpose of the PAM configuration files. Note that this man page discusses both /etc/pam.conf and individual configuration files in the /etc/pam.d/ directory. By default, Fedora uses the individual configuration files in the / etc/pam.d/ directory, ignoring /etc/pam.conf even if it exists. • pam_console — Describes the purpose of the pam_console.so module. It also describes the appropriate syntax for an entry within a PAM configuration file. • console.apps — Describes the format and options available in the /etc/security/ console.apps configuration file, which defines which applications are accessible by the console user assigned by PAM. • console.perms — Describes the format and options available in the /etc/security/ console.perms configuration file, which specifies the console user permissions assigned by PAM. • pam_timestamp — Describes the pam_timestamp.so module. • /usr/share/doc/pam- — Contains a System Administrators' Guide, a Module Writers' Manual, and the Application Developers' Manual, as well as a copy of the PAM standard, DCE-RFC 86.0, where is the version number of PAM. • /usr/share/doc/pam-/txts/README.pam_timestamp — Contains information about the pam_timestamp.so PAM module, where is the version number of PAM.

2.4.8.2. Useful PAM Websites • http://www.kernel.org/pub/linux/libs/pam/ — The primary distribution website for the Linux-PAM project, containing information on various PAM modules, a FAQ, and additional PAM documentation.

Note The documentation in the above website is for the last released upstream version of PAM and might not be 100% accurate for the PAM version included in Fedora.

2.5. TCP Wrappers and xinetd Controlling access to network services is one of the most important security tasks facing a server administrator. Fedora provides several tools for this purpose. For example, an iptables-based firewall filters out unwelcome network packets within the kernel's network stack. For network services that utilize it, TCP Wrappers add an additional layer of protection by defining which hosts are or are not allowed to connect to "wrapped" network services. One such wrapped network service is the xinetd super server. This service is called a super server because it controls connections to a subset of network services and further refines access control.

68

TCP Wrappers

Figure 2.9, “Access Control to Network Services” is a basic illustration of how these tools work together to protect network services.

Figure 2.9. Access Control to Network Services This chapter focuses on the role of TCP Wrappers and xinetd in controlling access to network services and reviews how these tools can be used to enhance both logging and utilization management. Refer to Section 2.9, “IPTables” for information about using firewalls with iptables.

2.5.1. TCP Wrappers The TCP Wrappers package (tcp_wrappers) is installed by default and provides host-based access control to network services. The most important component within the package is the /usr/lib/ libwrap.a library. In general terms, a TCP-wrapped service is one that has been compiled against the libwrap.a library. When a connection attempt is made to a TCP-wrapped service, the service first references the host's access files (/etc/hosts.allow and /etc/hosts.deny) to determine whether or not the client is allowed to connect. In most cases, it then uses the syslog daemon (syslogd) to write the name of the requesting client and the requested service to /var/log/secure or /var/log/messages. If a client is allowed to connect, TCP Wrappers release control of the connection to the requested service and take no further part in the communication between the client and the server.

69


In addition to access control and logging, TCP Wrappers can execute commands to interact with the client before denying or releasing control of the connection to the requested network service. Because TCP Wrappers are a valuable addition to any server administrator's arsenal of security tools, most network services within Fedora are linked to the libwrap.a library. Some such applications include /usr/sbin/sshd, /usr/sbin/sendmail, and /usr/sbin/xinetd.

Note To determine if a network service binary is linked to libwrap.a, type the following command as the root user: ldd | grep libwrap

Replace with the name of the network service binary. If the command returns straight to the prompt with no output, then the network service is not linked to libwrap.a. The following example indicates that /usr/sbin/sshd is linked to libwrap.a: [root@myServer ~]# ldd /usr/sbin/sshd | grep libwrap libwrap.so.0 => /lib/libwrap.so.0 (0x00655000) [root@myServer ~]#

2.5.1.1. Advantages of TCP Wrappers TCP Wrappers provide the following advantages over other network service control techniques: • Transparency to both the client and the wrapped network service — Both the connecting client and the wrapped network service are unaware that TCP Wrappers are in use. Legitimate users are logged and connected to the requested service while connections from banned clients fail. • Centralized management of multiple protocols — TCP Wrappers operate separately from the network services they protect, allowing many server applications to share a common set of access control configuration files, making for simpler management.

2.5.2. TCP Wrappers Configuration Files To determine if a client is allowed to connect to a service, TCP Wrappers reference the following two files, which are commonly referred to as hosts access files: • /etc/hosts.allow • /etc/hosts.deny When a TCP-wrapped service receives a client request, it performs the following steps: 1. It references /etc/hosts.allow. — The TCP-wrapped service sequentially parses the /etc/ hosts.allow file and applies the first rule specified for that service. If it finds a matching rule, it allows the connection. If not, it moves on to the next step.

70

TCP Wrappers Configuration Files

2. It references /etc/hosts.deny. — The TCP-wrapped service sequentially parses the /etc/ hosts.deny file. If it finds a matching rule, it denies the connection. If not, it grants access to the service. The following are important points to consider when using TCP Wrappers to protect network services: • Because access rules in hosts.allow are applied first, they take precedence over rules specified in hosts.deny. Therefore, if access to a service is allowed in hosts.allow, a rule denying access to that same service in hosts.deny is ignored. • The rules in each file are read from the top down and the first matching rule for a given service is the only one applied. The order of the rules is extremely important. • If no rules for the service are found in either file, or if neither file exists, access to the service is granted. • TCP-wrapped services do not cache the rules from the hosts access files, so any changes to hosts.allow or hosts.deny take effect immediately, without restarting network services.

Warning If the last line of a hosts access file is not a newline character (created by pressing the Enter key), the last rule in the file fails and an error is logged to either / var/log/messages or /var/log/secure. This is also the case for a rule that spans multiple lines without using the backslash character. The following example illustrates the relevant portion of a log message for a rule failure due to either of these circumstances: warning: /etc/hosts.allow, line 20: missing newline or line too long

2.5.2.1. Formatting Access Rules The format for both /etc/hosts.allow and /etc/hosts.deny is identical. Each rule must be on its own line. Blank lines or lines that start with a hash (#) are ignored. Each rule uses the following basic format to control access to network services: : [: : : ...]

• — A comma-separated list of process names (not service names) or the ALL wildcard. The daemon list also accepts operators (refer to Section 2.5.2.1.4, “Operators”) to allow greater flexibility. • — A comma-separated list of hostnames, host IP addresses, special patterns, or wildcards which identify the hosts affected by the rule. The client list also accepts operators listed in Section 2.5.2.1.4, “Operators” to allow greater flexibility. • — An optional action or colon-separated list of actions performed when the rule is triggered. Option fields support expansions, launch shell commands, allow or deny access, and alter logging behavior.

71


Note More information on the specialist terms above can be found elsewhere in this Guide: • Section 2.5.2.1.1, “Wildcards” • Section 2.5.2.1.2, “Patterns” • Section 2.5.2.2.4, “Expansions” • Section 2.5.2.2, “Option Fields” The following is a basic sample hosts access rule: vsftpd : .example.com

This rule instructs TCP Wrappers to watch for connections to the FTP daemon (vsftpd) from any host in the example.com domain. If this rule appears in hosts.allow, the connection is accepted. If this rule appears in hosts.deny, the connection is rejected. The next sample hosts access rule is more complex and uses two option fields: sshd : .example.com

\ : spawn /bin/echo `/bin/date` access denied>>/var/log/sshd.log \ : deny

Note that each option field is preceded by the backslash (\). Use of the backslash prevents failure of the rule due to length. This sample rule states that if a connection to the SSH daemon (sshd) is attempted from a host in the example.com domain, execute the echo command to append the attempt to a special log file, and deny the connection. Because the optional deny directive is used, this line denies access even if it appears in the hosts.allow file. Refer to Section 2.5.2.2, “Option Fields” for a more detailed look at available options.

2.5.2.1.1. Wildcards Wildcards allow TCP Wrappers to more easily match groups of daemons or hosts. They are used most frequently in the client list field of access rules. The following wildcards are available: • ALL — Matches everything. It can be used for both the daemon list and the client list. • LOCAL — Matches any host that does not contain a period (.), such as localhost. • KNOWN — Matches any host where the hostname and host address are known or where the user is known. • UNKNOWN — Matches any host where the hostname or host address are unknown or where the user is unknown. • PARANOID — Matches any host where the hostname does not match the host address.

72


Important The KNOWN, UNKNOWN, and PARANOID wildcards should be used with care, because they rely on functioning DNS server for correct operation. Any disruption to name resolution may prevent legitimate users from gaining access to a service.

2.5.2.1.2. Patterns Patterns can be used in the client field of access rules to more precisely specify groups of client hosts. The following is a list of common patterns for entries in the client field: • Hostname beginning with a period (.) — Placing a period at the beginning of a hostname matches all hosts sharing the listed components of the name. The following example applies to any host within the example.com domain: ALL : .example.com

• IP address ending with a period (.) — Placing a period at the end of an IP address matches all hosts sharing the initial numeric groups of an IP address. The following example applies to any host within the 192.168.x.x network: ALL : 192.168.

• IP address/netmask pair — Netmask expressions can also be used as a pattern to control access to a particular group of IP addresses. The following example applies to any host with an address range of 192.168.0.0 through 192.168.1.255: ALL : 192.168.0.0/255.255.254.0

Important When working in the IPv4 address space, the address/prefix length (prefixlen) pair declarations (CIDR notation) are not supported. Only IPv6 rules can use this format. • [IPv6 address]/prefixlen pair — [net]/prefixlen pairs can also be used as a pattern to control access to a particular group of IPv6 addresses. The following example would apply to any host with an address range of 3ffe:505:2:1:: through 3ffe:505:2:1:ffff:ffff:ffff:ffff: ALL : [3ffe:505:2:1::]/64

• The asterisk (*) — Asterisks can be used to match entire groups of hostnames or IP addresses, as long as they are not mixed in a client list containing other types of patterns. The following example would apply to any host within the example.com domain: ALL : *.example.com

73


• The slash (/) — If a client list begins with a slash, it is treated as a file name. This is useful if rules specifying large numbers of hosts are necessary. The following example refers TCP Wrappers to the /etc/telnet.hosts file for all Telnet connections: in.telnetd : /etc/telnet.hosts

Other, lesser used, patterns are also accepted by TCP Wrappers. Refer to the hosts_access man 5 page for more information.

Warning Be very careful when using hostnames and domain names. Attackers can use a variety of tricks to circumvent accurate name resolution. In addition, disruption to DNS service prevents even authorized users from using network services. It is, therefore, best to use IP addresses whenever possible.

2.5.2.1.3. Portmap and TCP Wrappers Portmap's implementation of TCP Wrappers does not support host look-ups, which means portmap can not use hostnames to identify hosts. Consequently, access control rules for portmap in hosts.allow or hosts.deny must use IP addresses, or the keyword ALL, for specifying hosts. Changes to portmap access control rules may not take effect immediately. You may need to restart the portmap service. Widely used services, such as NIS and NFS, depend on portmap to operate, so be aware of these limitations.

2.5.2.1.4. Operators At present, access control rules accept one operator, EXCEPT. It can be used in both the daemon list and the client list of a rule. The EXCEPT operator allows specific exceptions to broader matches within the same rule. In the following example from a hosts.allow file, all example.com hosts are allowed to connect to all services except cracker.example.com: ALL: .example.com EXCEPT cracker.example.com

In another example from a hosts.allow file, clients from the 192.168.0.x network can use all services except for FTP: ALL EXCEPT vsftpd: 192.168.0.

Note Organizationally, it is often easier to avoid using EXCEPT operators. This allows other administrators to quickly scan the appropriate files to see what hosts are allowed or denied access to services, without having to sort through EXCEPT operators.

74


2.5.2.2. Option Fields In addition to basic rules that allow and deny access, the Fedora implementation of TCP Wrappers supports extensions to the access control language through option fields. By using option fields in hosts access rules, administrators can accomplish a variety of tasks such as altering log behavior, consolidating access control, and launching shell commands.

2.5.2.2.1. Logging Option fields let administrators easily change the log facility and priority level for a rule by using the severity directive. In the following example, connections to the SSH daemon from any host in the example.com domain are logged to the default authpriv syslog facility (because no facility value is specified) with a priority of emerg: sshd : .example.com : severity emerg

It is also possible to specify a facility using the severity option. The following example logs any SSH connection attempts by hosts from the example.com domain to the local0 facility with a priority of alert: sshd : .example.com : severity local0.alert

Note In practice, this example does not work until the syslog daemon (syslogd) is configured to log to the local0 facility. Refer to the syslog.conf man page for information about configuring custom log facilities.

2.5.2.2.2. Access Control Option fields also allow administrators to explicitly allow or deny hosts in a single rule by adding the allow or deny directive as the final option. For example, the following two rules allow SSH connections from client-1.example.com, but deny connections from client-2.example.com: sshd : client-1.example.com : allow sshd : client-2.example.com : deny

By allowing access control on a per-rule basis, the option field allows administrators to consolidate all access rules into a single file: either hosts.allow or hosts.deny. Some administrators consider this an easier way of organizing access rules.

2.5.2.2.3. Shell Commands Option fields allow access rules to launch shell commands through the following two directives: • spawn — Launches a shell command as a child process. This directive can perform tasks like using /usr/sbin/safe_finger to get more information about the requesting client or create special log files using the echo command.

75


In the following example, clients attempting to access Telnet services from the example.com domain are quietly logged to a special file: in.telnetd : .example.com \ : spawn /bin/echo `/bin/date` from %h>>/var/log/telnet.log \ : allow

• twist — Replaces the requested service with the specified command. This directive is often used to set up traps for intruders (also called "honey pots"). It can also be used to send messages to connecting clients. The twist directive must occur at the end of the rule line. In the following example, clients attempting to access FTP services from the example.com domain are sent a message using the echo command: vsftpd : .example.com \ : twist /bin/echo "421 This domain has been black-listed. Access denied!"

For more information about shell command options, refer to the hosts_options man page.

2.5.2.2.4. Expansions Expansions, when used in conjunction with the spawn and twist directives, provide information about the client, server, and processes involved. The following is a list of supported expansions: • %a — Returns the client's IP address. • %A — Returns the server's IP address. • %c — Returns a variety of client information, such as the username and hostname, or the username and IP address. • %d — Returns the daemon process name. • %h — Returns the client's hostname (or IP address, if the hostname is unavailable). • %H — Returns the server's hostname (or IP address, if the hostname is unavailable). • %n — Returns the client's hostname. If unavailable, unknown is printed. If the client's hostname and host address do not match, paranoid is printed. • %N — Returns the server's hostname. If unavailable, unknown is printed. If the server's hostname and host address do not match, paranoid is printed. • %p — Returns the daemon's process ID. • %s —Returns various types of server information, such as the daemon process and the host or IP address of the server. • %u — Returns the client's username. If unavailable, unknown is printed. The following sample rule uses an expansion in conjunction with the spawn command to identify the client host in a customized log file.

76

xinetd

When connections to the SSH daemon (sshd) are attempted from a host in the example.com domain, execute the echo command to log the attempt, including the client hostname (by using the %h expansion), to a special file: sshd : .example.com \ : spawn /bin/echo `/bin/date` access denied to %h>>/var/log/sshd.log \ : deny

Similarly, expansions can be used to personalize messages back to the client. In the following example, clients attempting to access FTP services from the example.com domain are informed that they have been banned from the server: vsftpd : .example.com \ : twist /bin/echo "421 %h has been banned from this server!"

For a full explanation of available expansions, as well as additional access control options, refer to section 5 of the man pages for hosts_access (man 5 hosts_access) and the man page for hosts_options. Refer to Section 2.5.5, “Additional Resources” for more information about TCP Wrappers.

2.5.3. xinetd The xinetd daemon is a TCP-wrapped super service which controls access to a subset of popular network services, including FTP, IMAP, and Telnet. It also provides service-specific configuration options for access control, enhanced logging, binding, redirection, and resource utilization control. When a client attempts to connect to a network service controlled by xinetd, the super service receives the request and checks for any TCP Wrappers access control rules. If access is allowed, xinetd verifies that the connection is allowed under its own access rules for that service. It also checks that the service can have more resources allotted to it and that it is not in breach of any defined rules. If all these conditions are met (that is, access is allowed to the service; the service has not reached its resource limit; and the service is not in breach of any defined rule), xinetd then starts an instance of the requested service and passes control of the connection to it. After the connection has been established, xinetd takes no further part in the communication between the client and the server.

2.5.4. xinetd Configuration Files The configuration files for xinetd are as follows: • /etc/xinetd.conf — The global xinetd configuration file. • /etc/xinetd.d/ — The directory containing all service-specific files.

2.5.4.1. The /etc/xinetd.conf File The /etc/xinetd.conf file contains general configuration settings which affect every service under xinetd's control. It is read when the xinetd service is first started, so for configuration changes to take effect, you need to restart the xinetd service. The following is a sample /etc/xinetd.conf file: defaults

77


{ instances log_type log_on_success log_on_failure cps

= = = = =

60 SYSLOG authpriv HOST PID HOST 25 30

} includedir /etc/xinetd.d

These lines control the following aspects of xinetd: • instances — Specifies the maximum number of simultaneous requests that xinetd can process. • log_type — Configures xinetd to use the authpriv log facility, which writes log entries to the /var/log/secure file. Adding a directive such as FILE /var/log/xinetdlog would create a custom log file called xinetdlog in the /var/log/ directory. • log_on_success — Configures xinetd to log successful connection attempts. By default, the remote host's IP address and the process ID of the server processing the request are recorded. • log_on_failure — Configures xinetd to log failed connection attempts or if the connection was denied. • cps — Configures xinetd to allow no more than 25 connections per second to any given service. If this limit is exceeded, the service is retired for 30 seconds. • includedir /etc/xinetd.d/ — Includes options declared in the service-specific configuration files located in the /etc/xinetd.d/ directory. Refer to Section 2.5.4.2, “The /etc/xinetd.d/ Directory” for more information.

Note Often, both the log_on_success and log_on_failure settings in /etc/ xinetd.conf are further modified in the service-specific configuration files. More information may therefore appear in a given service's log file than the /etc/ xinetd.conf file may indicate. Refer to Section 2.5.4.3.1, “Logging Options” for further information.

2.5.4.2. The /etc/xinetd.d/ Directory The /etc/xinetd.d/ directory contains the configuration files for each service managed by xinetd and the names of the files correlate to the service. As with xinetd.conf, this directory is read only when the xinetd service is started. For any changes to take effect, the administrator must restart the xinetd service. The format of files in the /etc/xinetd.d/ directory use the same conventions as /etc/ xinetd.conf. The primary reason the configuration for each service is stored in a separate file is to make customization easier and less likely to affect other services. To gain an understanding of how these files are structured, consider the /etc/xinetd.d/krb5telnet file: service telnet { flags

78

= REUSE

xinetd Configuration Files

socket_type wait user server log_on_failure disable

= stream = no = root = /usr/kerberos/sbin/telnetd += USERID = yes

}

These lines control various aspects of the telnet service: • service — Specifies the service name, usually one of those listed in the /etc/services file. • flags — Sets any of a number of attributes for the connection. REUSE instructs xinetd to reuse the socket for a Telnet connection.

Note The REUSE flag is deprecated. All services now implicitly use the REUSE flag.

• socket_type — Sets the network socket type to stream. • wait — Specifies whether the service is single-threaded (yes) or multi-threaded (no). • user — Specifies which user ID the process runs under. • server — Specifies which binary executable to launch. • log_on_failure — Specifies logging parameters for log_on_failure in addition to those already defined in xinetd.conf. • disable — Specifies whether the service is disabled (yes) or enabled (no). Refer to the xinetd.conf man page for more information about these options and their usage.

2.5.4.3. Altering xinetd Configuration Files A range of directives is available for services protected by xinetd. This section highlights some of the more commonly used options.

2.5.4.3.1. Logging Options The following logging options are available for both /etc/xinetd.conf and the service-specific configuration files within the /etc/xinetd.d/ directory. The following is a list of some of the more commonly used logging options: • ATTEMPT — Logs the fact that a failed attempt was made (log_on_failure). • DURATION — Logs the length of time the service is used by a remote system (log_on_success). • EXIT — Logs the exit status or termination signal of the service (log_on_success). • HOST — Logs the remote host's IP address (log_on_failure and log_on_success). • PID — Logs the process ID of the server receiving the request (log_on_success).

79


• USERID — Logs the remote user using the method defined in RFC 1413 for all multi-threaded stream services (log_on_failure andlog_on_success). For a complete list of logging options, refer to the xinetd.conf man page.

2.5.4.3.2. Access Control Options Users of xinetd services can choose to use the TCP Wrappers hosts access rules, provide access control via the xinetd configuration files, or a mixture of both. Refer to Section 2.5.2, “TCP Wrappers Configuration Files” for more information about TCP Wrappers hosts access control files. This section discusses using xinetd to control access to services.

Note Unlike TCP Wrappers, changes to access control only take effect if the xinetd administrator restarts the xinetd service. Also, unlike TCP Wrappers, access control through xinetd only affects services controlled by xinetd. The xinetd hosts access control differs from the method used by TCP Wrappers. While TCP Wrappers places all of the access configuration within two files, /etc/hosts.allow and /etc/ hosts.deny, xinetd's access control is found in each service's configuration file in the /etc/ xinetd.d/ directory. The following hosts access options are supported by xinetd: • only_from — Allows only the specified hosts to use the service. • no_access — Blocks listed hosts from using the service. • access_times — Specifies the time range when a particular service may be used. The time range must be stated in 24-hour format notation, HH:MM-HH:MM. The only_from and no_access options can use a list of IP addresses or host names, or can specify an entire network. Like TCP Wrappers, combining xinetd access control with the enhanced logging configuration can increase security by blocking requests from banned hosts while verbosely recording each connection attempt. For example, the following /etc/xinetd.d/telnet file can be used to block Telnet access from a particular network group and restrict the overall time range that even allowed users can log in: service telnet { disable flags socket_type wait user server log_on_failure no_access log_on_success access_times }

80

= no = REUSE = stream = no = root = /usr/kerberos/sbin/telnetd += USERID = 172.16.45.0/24 += PID HOST EXIT = 09:45-16:15

xinetd Configuration Files

In this example, when a client system from the 10.0.1.0/24 network, such as 10.0.1.2, tries to access the Telnet service, it receives the following message: Connection closed by foreign host.

In addition, their login attempts are logged in /var/log/messages as follows: Sep Sep Sep

7 14:58:33 localhost xinetd[5285]: FAIL: telnet address from=172.16.45.107 7 14:58:33 localhost xinetd[5283]: START: telnet pid=5285 from=172.16.45.107 7 14:58:33 localhost xinetd[5283]: EXIT: telnet status=0 pid=5285 duration=0(sec)

When using TCP Wrappers in conjunction with xinetd access controls, it is important to understand the relationship between the two access control mechanisms. The following is the sequence of events followed by xinetd when a client requests a connection: 1. The xinetd daemon accesses the TCP Wrappers hosts access rules using a libwrap.a library call. If a deny rule matches the client, the connection is dropped. If an allow rule matches the client, the connection is passed to xinetd. 2. The xinetd daemon checks its own access control rules both for the xinetd service and the requested service. If a deny rule matches the client, the connection is dropped. Otherwise, xinetd starts an instance of the requested service and passes control of the connection to that service.

Important Care should be taken when using TCP Wrappers access controls in conjunction with xinetd access controls. Misconfiguration can cause undesirable effects.

2.5.4.3.3. Binding and Redirection Options The service configuration files for xinetd support binding the service to an IP address and redirecting incoming requests for that service to another IP address, hostname, or port. Binding is controlled with the bind option in the service-specific configuration files and links the service to one IP address on the system. When this is configured, the bind option only allows requests to the correct IP address to access the service. You can use this method to bind different services to different network interfaces based on requirements. This is particularly useful for systems with multiple network adapters or with multiple IP addresses. On such a system, insecure services (for example, Telnet), can be configured to listen only on the interface connected to a private network and not to the interface connected to the Internet. The redirect option accepts an IP address or hostname followed by a port number. It configures the service to redirect any requests for this service to the specified host and port number. This feature can be used to point to another port number on the same system, redirect the request to a different IP address on the same machine, shift the request to a totally different system and port number, or any combination of these options. A user connecting to a certain service on a system may therefore be rerouted to another system without disruption. The xinetd daemon is able to accomplish this redirection by spawning a process that stays alive for the duration of the connection between the requesting client machine and the host actually providing the service, transferring data between the two systems.

81


The advantages of the bind and redirect options are most clearly evident when they are used together. By binding a service to a particular IP address on a system and then redirecting requests for this service to a second machine that only the first machine can see, an internal system can be used to provide services for a totally different network. Alternatively, these options can be used to limit the exposure of a particular service on a multi-homed machine to a known IP address, as well as redirect any requests for that service to another machine especially configured for that purpose. For example, consider a system that is used as a firewall with this setting for its Telnet service: service telnet { socket_type = stream wait = no server = /usr/kerberos/sbin/telnetd log_on_success += DURATION USERID log_on_failure += USERID bind = 123.123.123.123 redirect = 10.0.1.13 23 }

The bind and redirect options in this file ensure that the Telnet service on the machine is bound to the external IP address (123.123.123.123), the one facing the Internet. In addition, any requests for Telnet service sent to 123.123.123.123 are redirected via a second network adapter to an internal IP address (10.0.1.13) that only the firewall and internal systems can access. The firewall then sends the communication between the two systems, and the connecting system thinks it is connected to 123.123.123.123 when it is actually connected to a different machine. This feature is particularly useful for users with broadband connections and only one fixed IP address. When using Network Address Translation (NAT), the systems behind the gateway machine, which are using internal-only IP addresses, are not available from outside the gateway system. However, when certain services controlled by xinetd are configured with the bind and redirect options, the gateway machine can act as a proxy between outside systems and a particular internal machine configured to provide the service. In addition, the various xinetd access control and logging options are also available for additional protection.

2.5.4.3.4. Resource Management Options The xinetd daemon can add a basic level of protection from Denial of Service (DoS) attacks. The following is a list of directives which can aid in limiting the effectiveness of such attacks: • per_source — Defines the maximum number of instances for a service per source IP address. It accepts only integers as an argument and can be used in both xinetd.conf and in the servicespecific configuration files in the xinetd.d/ directory. • cps — Defines the maximum number of connections per second. This directive takes two integer arguments separated by white space. The first argument is the maximum number of connections allowed to the service per second. The second argument is the number of seconds that xinetd must wait before re-enabling the service. It accepts only integers as arguments and can be used in either the xinetd.conf file or the service-specific configuration files in the xinetd.d/ directory. • max_load — Defines the CPU usage or load average threshold for a service. It accepts a floating point number argument. The load average is a rough measure of how many processes are active at a given time. See the uptime, who, and procinfo commands for more information about load average.

82


There are more resource management options available for xinetd. Refer to the xinetd.conf man page for more information.

2.5.5. Additional Resources More information about TCP Wrappers and xinetd is available from system documentation and on the Internet.

2.5.5.1. Installed TCP Wrappers Documentation The documentation on your system is a good place to start looking for additional configuration options for TCP Wrappers, xinetd, and access control. • /usr/share/doc/tcp_wrappers-/ — This directory contains a README file that discusses how TCP Wrappers work and the various hostname and host address spoofing risks that exist. • /usr/share/doc/xinetd-/ — This directory contains a README file that discusses aspects of access control and a sample.conf file with various ideas for modifying service-specific configuration files in the /etc/xinetd.d/ directory. • TCP Wrappers and xinetd-related man pages — A number of man pages exist for the various applications and configuration files involved with TCP Wrappers and xinetd. The following are some of the more important man pages: Server Applications • man xinetd — The man page for xinetd. Configuration Files • man 5 hosts_access — The man page for the TCP Wrappers hosts access control files. • man hosts_options — The man page for the TCP Wrappers options fields. • man xinetd.conf — The man page listing xinetd configuration options.

2.5.5.2. Useful TCP Wrappers Websites 4

• http://www.xinetd.org/ — The home of xinetd, containing sample configuration files, a full listing of features, and an informative FAQ. • http://www.docstoc.com/docs/2133633/An-Unofficial-Xinetd-Tutorial — A thorough tutorial that discusses many different ways to optimize default xinetd configuration files to meet specific security goals.

2.5.5.3. Related Books • Hacking Linux Exposed by Brian Hatch, James Lee, and George Kurtz; Osbourne/McGraw-Hill — An excellent security resource with information about TCP Wrappers and xinetd.

2.6. Kerberos System security and integrity within a network can be unwieldy. It can occupy the time of several administrators just to keep track of what services are being run on a network and the manner in which these services are used.

83


Further, authenticating users to network services can prove dangerous when the method used by the protocol is inherently insecure, as evidenced by the transfer of unencrypted passwords over a network using the traditional FTP and Telnet protocols. Kerberos is a way to eliminate the need for protocols that allow unsafe methods of authentication, thereby enhancing overall network security.

2.6.1. What is Kerberos? Kerberos is a network authentication protocol created by MIT, and uses symmetric-key cryptography to authenticate users to network services, which means passwords are never actually sent over the network.

5

Consequently, when users authenticate to network services using Kerberos, unauthorized users attempting to gather passwords by monitoring network traffic are effectively thwarted.

2.6.1.1. Advantages of Kerberos Most conventional network services use password-based authentication schemes. Such schemes require a user to authenticate to a given network server by supplying their username and password. Unfortunately, the transmission of authentication information for many services is unencrypted. For such a scheme to be secure, the network has to be inaccessible to outsiders, and all computers and users on the network must be trusted and trustworthy. Even if this is the case, a network that is connected to the Internet can no longer be assumed to be secure. Any attacker who gains access to the network can use a simple packet analyzer, also known as a packet sniffer, to intercept usernames and passwords, compromising user accounts and the integrity of the entire security infrastructure. The primary design goal of Kerberos is to eliminate the transmission of unencrypted passwords across the network. If used properly, Kerberos effectively eliminates the threat that packet sniffers would otherwise pose on a network.

2.6.1.2. Disadvantages of Kerberos Although Kerberos removes a common and severe security threat, it may be difficult to implement for a variety of reasons: • Migrating user passwords from a standard UNIX password database, such as /etc/passwd or /etc/shadow, to a Kerberos password database can be tedious, as there is no automated mechanism to perform this task. Refer to Question 2.23 in the online Kerberos FAQ: 6

http://www.nrl.navy.mil/CCS/people/kenh/kerberos-faq.html

• Kerberos has only partial compatibility with the Pluggable Authentication Modules (PAM) system used by most Fedora servers. Refer to Section 2.6.4, “Kerberos and PAM” for more information about this issue. • Kerberos assumes that each user is trusted but is using an untrusted host on an untrusted network. Its primary goal is to prevent unencrypted passwords from being transmitted across that network. However, if anyone other than the proper user has access to the one host that issues tickets used A system where both the client and the server share a common key that is used to encrypt and decrypt network communication.

84

Kerberos Terminology

for authentication — called the key distribution center (KDC) — the entire Kerberos authentication system is at risk. • For an application to use Kerberos, its source must be modified to make the appropriate calls into the Kerberos libraries. Applications modified in this way are considered to be Kerberos-aware, or kerberized. For some applications, this can be quite problematic due to the size of the application or its design. For other incompatible applications, changes must be made to the way in which the server and client communicate. Again, this may require extensive programming. Closed-source applications that do not have Kerberos support by default are often the most problematic. • Kerberos is an all-or-nothing solution. If Kerberos is used on the network, any unencrypted passwords transferred to a non-Kerberos aware service is at risk. Thus, the network gains no benefit from the use of Kerberos. To secure a network with Kerberos, one must either use Kerberosaware versions of all client/server applications that transmit passwords unencrypted, or not use any such client/server applications at all.

2.6.2. Kerberos Terminology Kerberos has its own terminology to define various aspects of the service. Before learning how Kerberos works, it is important to learn the following terms. authentication server (AS) A server that issues tickets for a desired service which are in turn given to users for access to the service. The AS responds to requests from clients who do not have or do not send credentials with a request. It is usually used to gain access to the ticket-granting server (TGS) service by issuing a ticket-granting ticket (TGT). The AS usually runs on the same host as the key distribution center (KDC). ciphertext Encrypted data. client An entity on the network (a user, a host, or an application) that can receive a ticket from Kerberos. credentials A temporary set of electronic credentials that verify the identity of a client for a particular service. Also called a ticket. credential cache or ticket file A file which contains the keys for encrypting communications between a user and various network services. Kerberos 5 supports a framework for using other cache types, such as shared memory, but files are more thoroughly supported. crypt hash A one-way hash used to authenticate users. These are more secure than using unencrypted data, but they are still relatively easy to decrypt for an experienced cracker. GSS-API The Generic Security Service Application Program Interface (defined in RFC-2743 published by The Internet Engineering Task Force) is a set of functions which provide security services. This API is used by clients and services to authenticate to each other without either program having specific knowledge of the underlying mechanism. If a network service (such as cyrus-IMAP) uses GSS-API, it can authenticate using Kerberos.

85


hash Also known as a hash value. A value generated by passing a string through a hash function. These values are typically used to ensure that transmitted data has not been tampered with. hash function A way of generating a digital "fingerprint" from input data. These functions rearrange, transpose or otherwise alter data to produce a hash value. key Data used when encrypting or decrypting other data. Encrypted data cannot be decrypted without the proper key or extremely good fortune on the part of the cracker. key distribution center (KDC) A service that issues Kerberos tickets, and which usually run on the same host as the ticketgranting server (TGS). keytab (or key table) A file that includes an unencrypted list of principals and their keys. Servers retrieve the keys they need from keytab files instead of using kinit. The default keytab file is /etc/krb5.keytab. The KDC administration server, /usr/kerberos/sbin/kadmind, is the only service that uses any other file (it uses /var/kerberos/krb5kdc/kadm5.keytab). kinit The kinit command allows a principal who has already logged in to obtain and cache the initial ticket-granting ticket (TGT). Refer to the kinit man page for more information. principal (or principal name) The principal is the unique name of a user or service allowed to authenticate using Kerberos. A principal follows the form root[/instance]@REALM. For a typical user, the root is the same as their login ID. The instance is optional. If the principal has an instance, it is separated from the root with a forward slash ("/"). An empty string ("") is considered a valid instance (which differs from the default NULL instance), but using it can be confusing. All principals in a realm have their own key, which for users is derived from a password or is randomly set for services. realm A network that uses Kerberos, composed of one or more servers called KDCs and a potentially large number of clients. service A program accessed over the network. ticket A temporary set of electronic credentials that verify the identity of a client for a particular service. Also called credentials. ticket-granting server (TGS) A server that issues tickets for a desired service which are in turn given to users for access to the service. The TGS usually runs on the same host as the KDC. ticket-granting ticket (TGT) A special ticket that allows the client to obtain additional tickets without applying for them from the KDC.

86

How Kerberos Works

unencrypted password A plain text, human-readable password.

2.6.3. How Kerberos Works Kerberos differs from username/password authentication methods. Instead of authenticating each user to each network service, Kerberos uses symmetric encryption and a trusted third party (a KDC), to authenticate users to a suite of network services. When a user authenticates to the KDC, the KDC sends a ticket specific to that session back to the user's machine, and any Kerberos-aware services look for the ticket on the user's machine rather than requiring the user to authenticate using a password. When a user on a Kerberos-aware network logs in to their workstation, their principal is sent to the KDC as part of a request for a TGT from the Authentication Server. This request can be sent by the log-in program so that it is transparent to the user, or can be sent by the kinit program after the user logs in. The KDC then checks for the principal in its database. If the principal is found, the KDC creates a TGT, which is encrypted using the user's key and returned to that user. The login or kinit program on the client then decrypts the TGT using the user's key, which it computes from the user's password. The user's key is used only on the client machine and is not transmitted over the network. The TGT is set to expire after a certain period of time (usually ten to twenty-four hours) and is stored in the client machine's credentials cache. An expiration time is set so that a compromised TGT is of use to an attacker for only a short period of time. After the TGT has been issued, the user does not have to re-enter their password until the TGT expires or until they log out and log in again. Whenever the user needs access to a network service, the client software uses the TGT to request a new ticket for that specific service from the TGS. The service ticket is then used to authenticate the user to that service transparently.

Warning The Kerberos system can be compromised if a user on the network authenticates against a non-Kerberos aware service by transmitting a password in plain text. The use of non-Kerberos aware services is highly discouraged. Such services include Telnet and FTP. The use of other encrypted protocols, such as SSH or SSL-secured services, however, is preferred, although not ideal. This is only a broad overview of how Kerberos authentication works. Refer to Section 2.6.10, “Additional Resources” for links to more in-depth information.

87


Note Kerberos depends on the following network services to function correctly. • Approximate clock synchronization between the machines on the network. A clock synchronization program should be set up for the network, such as ntpd. Refer to /usr/share/doc/ntp-/index.html for details on setting up Network Time Protocol servers (where is the version number of the ntp package installed on your system). • Domain Name Service (DNS). You should ensure that the DNS entries and hosts on the network are all properly configured. Refer to the Kerberos V5 System Administrator's Guide in /usr/ share/doc/krb5-server- for more information (where is the version number of the krb5-server package installed on your system).

2.6.4. Kerberos and PAM Kerberos-aware services do not currently make use of Pluggable Authentication Modules (PAM) — these services bypass PAM completely. However, applications that use PAM can make use of Kerberos for authentication if the pam_krb5 module (provided in the pam_krb5 package) is installed. The pam_krb5 package contains sample configuration files that allow services such as login and gdm to authenticate users as well as obtain initial credentials using their passwords. If access to network servers is always performed using Kerberos-aware services or services that use GSS-API, such as IMAP, the network can be considered reasonably safe.

Important Administrators should be careful not to allow users to authenticate to most network services using Kerberos passwords. Many protocols used by these services do not encrypt the password before sending it over the network, destroying the benefits of the Kerberos system. For example, users should not be allowed to authenticate to Telnet services with the same password they use for Kerberos authentication.

2.6.5. Configuring a Kerberos 5 Server When setting up Kerberos, install the KDC first. If it is necessary to set up slave servers, install the master first. To configure the first Kerberos KDC, follow these steps: 1.

88

Ensure that time synchronization and DNS are functioning correctly on all client and server machines before configuring Kerberos. Pay particular attention to time synchronization between the Kerberos server and its clients. If the time difference between the server and client is greater than five minutes (this is configurable in Kerberos 5), Kerberos clients can not authenticate to the server. This time synchronization is necessary to prevent an attacker from using an old Kerberos ticket to masquerade as a valid user.

Configuring a Kerberos 5 Server

It is advisable to set up a Network Time Protocol (NTP) compatible client/server network even if Kerberos is not being used. Fedora includes the ntp package for this purpose. Refer to /usr/ share/doc/ntp-/index.html (where is the version number of the ntp package installed on your system) for details about how to set up Network Time Protocol servers, and http://www.ntp.org for more information about NTP. 2.

Install the krb5-libs, krb5-server, and krb5-workstation packages on the dedicated machine which runs the KDC. This machine needs to be very secure — if possible, it should not run any services other than the KDC.

3.

Edit the /etc/krb5.conf and /var/kerberos/krb5kdc/kdc.conf configuration files to reflect the realm name and domain-to-realm mappings. A simple realm can be constructed by replacing instances of EXAMPLE.COM and example.com with the correct domain name — being certain to keep uppercase and lowercase names in the correct format — and by changing the KDC from kerberos.example.com to the name of the Kerberos server. By convention, all realm names are uppercase and all DNS hostnames and domain names are lowercase. For full details about the formats of these configuration files, refer to their respective man pages.

4.

Create the database using the kdb5_util utility from a shell prompt: /usr/kerberos/sbin/kdb5_util create -s

The create command creates the database that stores keys for the Kerberos realm. The -s switch forces creation of a stash file in which the master server key is stored. If no stash file is present from which to read the key, the Kerberos server (krb5kdc) prompts the user for the master server password (which can be used to regenerate the key) every time it starts. 5.

Edit the /var/kerberos/krb5kdc/kadm5.acl file. This file is used by kadmind to determine which principals have administrative access to the Kerberos database and their level of access. Most organizations can get by with a single line: */[email protected]

*

Most users are represented in the database by a single principal (with a NULL, or empty, instance, such as [email protected]). In this configuration, users with a second principal with an instance of admin (for example, joe/[email protected]) are able to wield full power over the realm's Kerberos database. After kadmind has been started on the server, any user can access its services by running kadmin on any of the clients or servers in the realm. However, only users listed in the kadm5.acl file can modify the database in any way, except for changing their own passwords.

Note The kadmin utility communicates with the kadmind server over the network, and uses Kerberos to handle authentication. Consequently, the first principal must already exist before connecting to the server over the network to administer it. Create the first principal with the kadmin.local command, which is specifically designed to be used on the same host as the KDC and does not use Kerberos for authentication.

89


Type the following kadmin.local command at the KDC terminal to create the first principal: /usr/kerberos/sbin/kadmin.local -q "addprinc username/admin"

6.

Start Kerberos using the following commands: /sbin/service krb5kdc start /sbin/service kadmin start /sbin/service krb524 start

7.

Add principals for the users using the addprinc command within kadmin. kadmin and kadmin.local are command line interfaces to the KDC. As such, many commands — such as addprinc — are available after launching the kadmin program. Refer to the kadmin man page for more information.

8.

Verify that the KDC is issuing tickets. First, run kinit to obtain a ticket and store it in a credential cache file. Next, use klist to view the list of credentials in the cache and use kdestroy to destroy the cache and the credentials it contains.

Note By default, kinit attempts to authenticate using the same system login username (not the Kerberos server). If that username does not correspond to a principal in the Kerberos database, kinit issues an error message. If that happens, supply kinit with the name of the correct principal as an argument on the command line (kinit ). Once these steps are completed, the Kerberos server should be up and running.

2.6.6. Configuring a Kerberos 5 Client Setting up a Kerberos 5 client is less involved than setting up a server. At a minimum, install the client packages and provide each client with a valid krb5.conf configuration file. While ssh and slogin are the preferred method of remotely logging in to client systems, Kerberized versions of rsh and rlogin are still available, though deploying them requires that a few more configuration changes be made. 1.

Be sure that time synchronization is in place between the Kerberos client and the KDC. Refer to Section 2.6.5, “Configuring a Kerberos 5 Server” for more information. In addition, verify that DNS is working properly on the Kerberos client before configuring the Kerberos client programs.

2.

Install the krb5-libs and krb5-workstation packages on all of the client machines. Supply a valid /etc/krb5.conf file for each client (usually this can be the same krb5.conf file used by the KDC).

3.

Before a workstation in the realm can use Kerberos to authenticate users who connect using ssh or Kerberized rsh or rlogin, it must have its own host principal in the Kerberos database. The sshd, kshd, and klogind server programs all need access to the keys for the host service's principal. Additionally, in order to use the kerberized rsh and rlogin services, that workstation must have the xinetd package installed.

90

Domain-to-Realm Mapping

Using kadmin, add a host principal for the workstation on the KDC. The instance in this case is the hostname of the workstation. Use the -randkey option for the kadmin's addprinc command to create the principal and assign it a random key: addprinc -randkey host/blah.example.com

Now that the principal has been created, keys can be extracted for the workstation by running kadmin on the workstation itself, and using the ktadd command within kadmin: ktadd -k /etc/krb5.keytab host/blah.example.com

4.

To use other kerberized network services, they must first be started. Below is a list of some common kerberized services and instructions about enabling them: • ssh — OpenSSH uses GSS-API to authenticate users to servers if the client's and server's configuration both have GSSAPIAuthentication enabled. If the client also has GSSAPIDelegateCredentials enabled, the user's credentials are made available on the remote system. • rsh and rlogin — To use the kerberized versions of rsh and rlogin, enable klogin, eklogin, and kshell. • Telnet — To use kerberized Telnet, krb5-telnet must be enabled. • FTP — To provide FTP access, create and extract a key for the principal with a root of ftp. Be certain to set the instance to the fully qualified hostname of the FTP server, then enable gssftp. • IMAP — To use a kerberized IMAP server, the cyrus-imap package uses Kerberos 5 if it also has the cyrus-sasl-gssapi package installed. The cyrus-sasl-gssapi package contains the Cyrus SASL plugins which support GSS-API authentication. Cyrus IMAP should function properly with Kerberos as long as the cyrus user is able to find the proper key in / etc/krb5.keytab, and the root for the principal is set to imap (created with kadmin). An alternative to cyrus-imap can be found in the dovecot package, which is also included in Fedora. This package contains an IMAP server but does not, to date, support GSS-API and Kerberos. • CVS — To use a kerberized CVS server, gserver uses a principal with a root of cvs and is otherwise identical to the CVS pserver.

2.6.7. Domain-to-Realm Mapping When a client attempts to access a service running on a particular server, it knows the name of the service (host) and the name of the server (foo.example.com), but because more than one realm may be deployed on your network, it must guess at the name of the realm in which the service resides. By default, the name of the realm is taken to be the DNS domain name of the server, upper-cased. foo.example.org → EXAMPLE.ORG foo.example.com → EXAMPLE.COM foo.hq.example.com → HQ.EXAMPLE.COM

91


In some configurations, this will be sufficient, but in others, the realm name which is derived will be the name of a non-existant realm. In these cases, the mapping from the server's DNS domain name to the name of its realm must be specified in the domain_realm section of the client system's krb5.conf. For example: [domain_realm] .example.com = EXAMPLE.COM example.com = EXAMPLE.COM

The above configuration specifies two mappings. The first mapping specifies that any system in the "example.com" DNS domain belongs to the EXAMPLE.COM realm. The second specifies that a system with the exact name "example.com" is also in the realm. (The distinction between a domain and a specific host is marked by the presence or lack of an initial ".".) The mapping can also be stored directly in DNS.

2.6.8. Setting Up Secondary KDCs For a number of reasons, you may choose to run multiple KDCs for a given realm. In this scenario, one KDC (the master KDC) keeps a writable copy of the realm database and runs kadmind (it is also your realm's admin server), and one or more KDCs (slave KDCs) keep read-only copies of the database and run kpropd. The master-slave propagation procedure entails the master KDC dumping its database to a temporary dump file and then transmitting that file to each of its slaves, which then overwrite their previouslyreceived read-only copies of the database with the contents of the dump file. To set up a slave KDC, first ensure that the master KDC's krb5.conf and kdc.conf files are copied to the slave KDC. Start kadmin.local from a root shell on the master KDC and use its add_principal command to create a new entry for the master KDC's host service, and then use its ktadd command to simultaneously set a random key for the service and store the random key in the master's default keytab file. This key will be used by the kprop command to authenticate to the slave servers. You will only need to do this once, regardless of how many slave servers you install. # kadmin.local -r EXAMPLE.COM Authenticating as principal root/[email protected] with password. kadmin: add_principal -randkey host/masterkdc.example.com Principal "host/host/[email protected]" created. kadmin: ktadd host/masterkdc.example.com Entry for principal host/masterkdc.example.com with kvno 3, encryption type Triple DES cbc mode with HMAC/sha1 added to keytab WRFILE:/etc/krb5.keytab. Entry for principal host/masterkdc.example.com with kvno 3, encryption type ArcFour with HMAC/ md5 added to keytab WRFILE:/etc/krb5.keytab. Entry for principal host/masterkdc.example.com with kvno 3, encryption type DES with HMAC/sha1 added to keytab WRFILE:/etc/krb5.keytab. Entry for principal host/masterkdc.example.com with kvno 3, encryption type DES cbc mode with RSA-MD5 added to keytab WRFILE:/etc/krb5.keytab.

92

Setting Up Secondary KDCs

kadmin: quit

Start kadmin from a root shell on the slave KDC and use its add_principal command to create a new entry for the slave KDC's host service, and then use kadmin's ktadd command to simultaneously set a random key for the service and store the random key in the slave's default keytab file. This key is used by the kpropd service when authenticating clients. # kadmin -p jimbo/[email protected] -r EXAMPLE.COM Authenticating as principal jimbo/[email protected] with password. Password for jimbo/[email protected]: kadmin: add_principal -randkey host/slavekdc.example.com Principal "host/[email protected]" created. kadmin: ktadd host/[email protected] Entry for principal host/slavekdc.example.com with kvno 3, encryption type Triple DES cbc mode with HMAC/sha1 added to keytab WRFILE:/etc/krb5.keytab. Entry for principal host/slavekdc.example.com with kvno 3, encryption type ArcFour with HMAC/ md5 added to keytab WRFILE:/etc/krb5.keytab. Entry for principal host/slavekdc.example.com with kvno 3, encryption type DES with HMAC/sha1 added to keytab WRFILE:/etc/krb5.keytab. Entry for principal host/slavekdc.example.com with kvno 3, encryption type DES cbc mode with RSA-MD5 added to keytab WRFILE:/etc/krb5.keytab. kadmin: quit

With its service key, the slave KDC could authenticate any client which would connect to it. Obviously, not all of them should be allowed to provide the slave's kprop service with a new realm database. To restrict access, the kprop service on the slave KDC will only accept updates from clients whose principal names are listed in /var/kerberos/krb5kdc/kpropd.acl. Add the master KDC's host service's name to that file. # echo host/[email protected] > /var/kerberos/krb5kdc/kpropd.acl

Once the slave KDC has obtained a copy of the database, it will also need the master key which was used to encrypt it. If your KDC database's master key is stored in a stash file on the master KDC (typically named /var/kerberos/krb5kdc/.k5.REALM, either copy it to the slave KDC using any available secure method, or create a dummy database and identical stash file on the slave KDC by running kdb5_util create -s (the dummy database will be overwritten by the first successful database propagation) and supplying the same password. Ensure that the slave KDC's firewall allows the master KDC to contact it using TCP on port 754 (krb5_prop), and start the kprop service. Then, double-check that the kadmin service is disabled. Now perform a manual database propagation test by dumping the realm database, on the master KDC, to the default data file which the kprop command will read (/var/kerberos/krb5kdc/ slave_datatrans), and then use the kprop command to transmit its contents to the slave KDC. # /usr/kerberos/sbin/kdb5_util dump /var/kerberos/krb5kdc/slave_datatrans# kprop slavekdc.example.com

93


Using kinit, verify that a client system whose krb5.conf lists only the slave KDC in its list of KDCs for your realm is now correctly able to obtain initial credentials from the slave KDC. That done, simply create a script which dumps the realm database and runs the kprop command to transmit the database to each slave KDC in turn, and configure the cron service to run the script periodically.

2.6.9. Setting Up Cross Realm Authentication Cross-realm authentication is the term which is used to describe situations in which clients (typically users) of one realm use Kerberos to authenticate to services (typically server processes running on a particular server system) which belong to a realm other than their own. For the simplest case, in order for a client of a realm named A.EXAMPLE.COM to access a service in the B.EXAMPLE.COM realm, both realms must share a key for a principal named krbtgt/ [email protected], and both keys must have the same key version number associated with them. To accomplish this, select a very strong password or passphrase, and create an entry for the principal in both realms using kadmin. # kadmin -r A.EXAMPLE.COM kadmin: add_principal krbtgt/[email protected] Enter password for principal "krbtgt/[email protected]": Re-enter password for principal "krbtgt/[email protected]": Principal "krbtgt/[email protected]" created. quit # kadmin -r B.EXAMPLE.COM kadmin: add_principal krbtgt/[email protected] Enter password for principal "krbtgt/[email protected]": Re-enter password for principal "krbtgt/[email protected]": Principal "krbtgt/[email protected]" created. quit

Use the get_principal command to verify that both entries have matching key version numbers (kvno values) and encryption types.

Dumping the Database Doesn't Do It Security-conscious administrators may attempt to use the add_principal command's -randkey option to assign a random key instead of a password, dump the new entry from the database of the first realm, and import it into the second. This will not work unless the master keys for the realm databases are identical, as the keys contained in a database dump are themselves encrypted using the master key. Clients in the A.EXAMPLE.COM realm are now able to authenticate to services in the B.EXAMPLE.COM realm. Put another way, the B.EXAMPLE.COM realm now trusts the A.EXAMPLE.COM realm, or phrased even more simply, B.EXAMPLE.COM now trusts A.EXAMPLE.COM. This brings us to an important point: cross-realm trust is unidirectional by default. The KDC for the B.EXAMPLE.COM realm may trust clients from the A.EXAMPLE.COM to authenticate to services in the B.EXAMPLE.COM realm, but the fact that it does has no effect on whether or not clients in the B.EXAMPLE.COM realm are trusted to authenticate to services in the A.EXAMPLE.COM realm. To establish trust in the other direction, both realms would need to share keys for the krbtgt/ [email protected] service (take note of the reversed in order of the two realms compared to the example above). If direct trust relationships were the only method for providing trust between realms, networks which contain multiple realms would be very difficult to set up. Luckily, cross-realm trust is transitive. If

94

Setting Up Cross Realm Authentication

clients from A.EXAMPLE.COM can authenticate to services in B.EXAMPLE.COM, and clients from B.EXAMPLE.COM can authenticate to services in C.EXAMPLE.COM, then clients in A.EXAMPLE.COM can also authenticate to services in C.EXAMPLE.COM, even if C.EXAMPLE.COM doesn't directly trust A.EXAMPLE.COM. This means that, on a network with multiple realms which all need to trust each other, making good choices about which trust relationships to set up can greatly reduce the amount of effort required. Now you face the more conventional problems: the client's system must be configured so that it can properly deduce the realm to which a particular service belongs, and it must be able to determine how to obtain credentials for services in that realm. First things first: the principal name for a service provided from a specific server system in a given realm typically looks like this: service/[email protected]

In this example, service is typically either the name of the protocol in use (other common values include ldap, imap, cvs, and HTTP) or host, server.example.com is the fully-qualified domain name of the system which runs the service, and EXAMPLE.COM is the name of the realm. To deduce the realm to which the service belongs, clients will most often consult DNS or the domain_realm section of /etc/krb5.conf to map either a hostname (server.example.com) or a DNS domain name (.example.com) to the name of a realm (EXAMPLE.COM). Having determined which to which realm a service belongs, a client then has to determine the set of realms which it needs to contact, and in which order it must contact them, to obtain credentials for use in authenticating to the service. This can be done in one of two ways. The default method, which requires no explicit configuration, is to give the realms names within a shared hierarchy. For an example, assume realms named A.EXAMPLE.COM, B.EXAMPLE.COM, and EXAMPLE.COM. When a client in the A.EXAMPLE.COM realm attempts to authenticate to a service in B.EXAMPLE.COM, it will, by default, first attempt to get credentials for the EXAMPLE.COM realm, and then to use those credentials to obtain credentials for use in the B.EXAMPLE.COM realm. The client in this scenario treats the realm name as one might treat a DNS name. It repeatedly strips off the components of its own realm's name to generate the names of realms which are "above" it in the hierarchy until it reaches a point which is also "above" the service's realm. At that point it begins prepending components of the service's realm name until it reaches the service's realm. Each realm which is involved in the process is another "hop". For example, using credentials in A.EXAMPLE.COM, authenticating to a service in B.EXAMPLE.COMA.EXAMPLE.COM → EXAMPLE.COM → B.EXAMPLE.COM • A.EXAMPLE.COM and EXAMPLE.COM share a key for krbtgt/[email protected] • EXAMPLE.COM and B.EXAMPLE.COM share a key for krbtgt/[email protected] Another example, using credentials in SITE1.SALES.EXAMPLE.COM, authenticating to a service in EVERYWHERE.EXAMPLE.COMSITE1.SALES.EXAMPLE.COM → SALES.EXAMPLE.COM → EXAMPLE.COM → EVERYWHERE.EXAMPLE.COM • SITE1.SALES.EXAMPLE.COM and SALES.EXAMPLE.COM share a key for krbtgt/ [email protected]

95


• SALES.EXAMPLE.COM and EXAMPLE.COM share a key for krbtgt/ [email protected] • EXAMPLE.COM and EVERYWHERE.EXAMPLE.COM share a key for krbtgt/ [email protected] Another example, this time using realm names whose names share no common suffix (DEVEL.EXAMPLE.COM and PROD.EXAMPLE.ORG DEVEL.EXAMPLE.COM → EXAMPLE.COM → COM → ORG → EXAMPLE.ORG → PROD.EXAMPLE.ORG • DEVEL.EXAMPLE.COM and EXAMPLE.COM share a key for krbtgt/ [email protected] • EXAMPLE.COM and COM share a key for krbtgt/[email protected] • COM and ORG share a key for krbtgt/ORG@COM • ORG and EXAMPLE.ORG share a key for krbtgt/EXAMPLE.ORG@ORG • EXAMPLE.ORG and PROD.EXAMPLE.ORG share a key for krbtgt/ [email protected] The more complicated, but also more flexible, method involves configuring the capaths section of / etc/krb5.conf, so that clients which have credentials for one realm will be able to look up which realm is next in the chain which will eventually lead to the being able to authenticate to servers. The format of the capaths section is relatively straightforward: each entry in the section is named after a realm in which a client might exist. Inside of that subsection, the set of intermediate realms from which the client must obtain credentials is listed as values of the key which corresponds to the realm in which a service might reside. If there are no intermediate realms, the value "." is used. Here's an example: [capaths] A.EXAMPLE.COM = { B.EXAMPLE.COM = . C.EXAMPLE.COM = B.EXAMPLE.COM D.EXAMPLE.COM = B.EXAMPLE.COM D.EXAMPLE.COM = C.EXAMPLE.COM }

In this example, clients in the A.EXAMPLE.COM realm can obtain cross-realm credentials for B.EXAMPLE.COM directly from the A.EXAMPLE.COM KDC. If those clients wish to contact a service in theC.EXAMPLE.COM realm, they will first need to obtain necessary credentials from the B.EXAMPLE.COM realm (this requires that krbtgt/ [email protected] exist), and then use those credentials to obtain credentials for use in the C.EXAMPLE.COM realm (using krbtgt/[email protected]). If those clients wish to contact a service in the D.EXAMPLE.COM realm, they will first need to obtain necessary credentials from the B.EXAMPLE.COM realm, and then credentials from the C.EXAMPLE.COM realm, before finally obtaining credentials for use with the D.EXAMPLE.COM realm.

96


Note Without a capath entry indicating otherwise, Kerberos assumes that cross-realm trust relationships form a hierarchy. Clients in the A.EXAMPLE.COM realm can obtain cross-realm credentials from B.EXAMPLE.COM realm directly. Without the "." indicating this, the client would instead attempt to use a hierarchical path, in this case: A.EXAMPLE.COM → EXAMPLE.COM → B.EXAMPLE.COM

2.6.10. Additional Resources For more information about Kerberos, refer to the following resources.

2.6.10.1. Installed Kerberos Documentation • The Kerberos V5 Installation Guide and the Kerberos V5 System Administrator's Guide in PostScript and HTML formats. These can be found in the /usr/share/doc/krb5-server-/ directory (where is the version number of the krb5-server package installed on your system). • The Kerberos V5 UNIX User's Guide in PostScript and HTML formats. These can be found in the /usr/share/doc/krb5-workstation-/ directory (where is the version number of the krb5-workstation package installed on your system). • Kerberos man pages — There are a number of man pages for the various applications and configuration files involved with a Kerberos implementation. The following is a list of some of the more important man pages. Client Applications • man kerberos — An introduction to the Kerberos system which describes how credentials work and provides recommendations for obtaining and destroying Kerberos tickets. The bottom of the man page references a number of related man pages. • man kinit — Describes how to use this command to obtain and cache a ticket-granting ticket. • man kdestroy — Describes how to use this command to destroy Kerberos credentials. • man klist — Describes how to use this command to list cached Kerberos credentials. Administrative Applications • man kadmin — Describes how to use this command to administer the Kerberos V5 database. • man kdb5_util — Describes how to use this command to create and perform low-level administrative functions on the Kerberos V5 database. Server Applications • man krb5kdc — Describes available command line options for the Kerberos V5 KDC.

97


• man kadmind — Describes available command line options for the Kerberos V5 administration server. Configuration Files • man krb5.conf — Describes the format and options available within the configuration file for the Kerberos V5 library. • man kdc.conf — Describes the format and options available within the configuration file for the Kerberos V5 AS and KDC.

2.6.10.2. Useful Kerberos Websites • http://web.mit.edu/kerberos/www/ — Kerberos: The Network Authentication Protocol webpage from MIT. • http://www.nrl.navy.mil/CCS/people/kenh/kerberos-faq.html — The Kerberos Frequently Asked Questions (FAQ). • ftp://athena-dist.mit.edu/pub/kerberos/doc/usenix.PS — The PostScript version of Kerberos: An Authentication Service for Open Network Systems by Jennifer G. Steiner, Clifford Neuman, and Jeffrey I. Schiller. This document is the original paper describing Kerberos. • http://web.mit.edu/kerberos/www/dialogue.html — Designing an Authentication System: a Dialogue in Four Scenes originally by Bill Bryant in 1988, modified by Theodore Ts'o in 1997. This document is a conversation between two developers who are thinking through the creation of a Kerberos-style authentication system. The conversational style of the discussion make this a good starting place for people who are completely unfamiliar with Kerberos. • http://www.ornl.gov/~jar/HowToKerb.html — How to Kerberize your site is a good reference for kerberizing a network. • http://www.networkcomputing.com/netdesign/kerb1.html — Kerberos Network Design Manual is a thorough overview of the Kerberos system.

2.7. Virtual Private Networks (VPNs) Organizations with several satellite offices often connect to each other with dedicated lines for efficiency and protection of sensitive data in transit. For example, many businesses use frame relay or Asynchronous Transfer Mode (ATM) lines as an end-to-end networking solution to link one office with others. This can be an expensive proposition, especially for small to medium sized businesses (SMBs) that want to expand without paying the high costs associated with enterprise-level, dedicated digital circuits. To address this need, Virtual Private Networks (VPNs) were developed. Following the same functional principles as dedicated circuits, VPNs allow for secured digital communication between two parties (or networks), creating a Wide Area Network (WAN) from existing Local Area Networks (LANs). Where it differs from frame relay or ATM is in its transport medium. VPNs transmit over IP using datagrams as the transport layer, making it a secure conduit through the Internet to an intended destination. Most free software VPN implementations incorporate open standard encryption methods to further mask data in transit. Some organizations employ hardware VPN solutions to augment security, while others use software or protocol-based implementations. Several vendors provide hardware VPN solutions, such as Cisco,

98

How Does a VPN Work?

Nortel, IBM, and Checkpoint. There is a free software-based VPN solution for Linux called FreeS/Wan that utilizes a standardized Internet Protocol Security (IPsec) implementation. These VPN solutions, irrespective of whether they are hardware or software based, act as specialized routers that exist between the IP connection from one office to another.

2.7.1. How Does a VPN Work? When a packet is transmitted from a client, it sends it through the VPN router or gateway, which adds an Authentication Header (AH) for routing and authentication. The data is then encrypted and, finally, enclosed with an Encapsulating Security Payload (ESP). This latter constitutes the decryption and handling instructions. The receiving VPN router strips the header information, decrypts the data, and routes it to its intended destination (either a workstation or other node on a network). Using a network-to-network connection, the receiving node on the local network receives the packets already decrypted and ready for processing. The encryption/decryption process in a network-to-network VPN connection is transparent to a local node. With such a heightened level of security, an attacker must not only intercept a packet, but decrypt the packet as well. Intruders who employ a man-in-the-middle attack between a server and client must also have access to at least one of the private keys for authenticating sessions. Because they employ several layers of authentication and encryption, VPNs are a secure and effective means of connecting multiple remote nodes to act as a unified intranet.

2.7.2. VPNs and Fedora Fedora provides various options in terms of implementing a software solution to securely connect to a WAN. Internet Protocol Security (IPsec) is the supported VPN implementation for Fedora, and sufficiently addresses the usability needs of organizations with branch offices or remote users.

2.7.3. IPsec Fedora supports IPsec for connecting remote hosts and networks to each other using a secure tunnel on a common carrier network such as the Internet. IPsec can be implemented using a host-to-host (one computer workstation to another) or network-to-network (one LAN/WAN to another) configuration. The IPsec implementation in Fedora uses Internet Key Exchange (IKE), a protocol implemented by the Internet Engineering Task Force (IETF), used for mutual authentication and secure associations between connecting systems.

2.7.4. Creating an IPsec Connection An IPsec connection is split into two logical phases. In phase 1, an IPsec node initializes the connection with the remote node or network. The remote node or network checks the requesting node's credentials and both parties negotiate the authentication method for the connection. On Fedora systems, an IPsec connection uses the pre-shared key method of IPsec node authentication. In a pre-shared key IPsec connection, both hosts must use the same key in order to move to Phase 2 of the IPsec connection. Phase 2 of the IPsec connection is where the Security Association (SA) is created between IPsec nodes. This phase establishes an SA database with configuration information, such as the encryption method, secret session key exchange parameters, and more. This phase manages the actual IPsec connection between remote nodes and networks.

99


The Fedora implementation of IPsec uses IKE for sharing keys between hosts across the Internet. The racoon keying daemon handles the IKE key distribution and exchange. Refer to the racoon man page for more information about this daemon.

2.7.5. IPsec Installation Implementing IPsec requires that the ipsec-tools RPM package be installed on all IPsec hosts (if using a host-to-host configuration) or routers (if using a network-to-network configuration). The RPM package contains essential libraries, daemons, and configuration files for setting up the IPsec connection, including: • /sbin/setkey — manipulates the key management and security attributes of IPsec in the kernel. This executable is controlled by the racoon key management daemon. Refer to the setkey(8) man page for more information. • /usr/sbin/racoon — the IKE key management daemon, used to manage and control security associations and key sharing between IPsec-connected systems. • /etc/racoon/racoon.conf — the racoon daemon configuration file used to configure various aspects of the IPsec connection, including authentication methods and encryption algorithms used in the connection. Refer to the racoon.conf(5) man page for a complete listing of available directives. To configure IPsec on Fedora, you can use the Network Administration Tool, or manually edit the networking and IPsec configuration files. • To connect two network-connected hosts via IPsec, refer to Section 2.7.6, “IPsec Host-to-Host Configuration”. • To connect one LAN/WAN to another via IPsec, refer to Section 2.7.7, “IPsec Network-to-Network Configuration”.

2.7.6. IPsec Host-to-Host Configuration IPsec can be configured to connect one desktop or workstation (host) to another using a host-tohost connection. This type of connection uses the network to which each host is connected to create a secure tunnel between each host. The requirements of a host-to-host connection are minimal, as is the configuration of IPsec on each host. The hosts need only a dedicated connection to a carrier network (such as the Internet) and Fedora to create the IPsec connection.

2.7.6.1. Host-to-Host Connection A host-to-host IPsec connection is an encrypted connection between two systems, both running IPsec with the same authentication key. With the IPsec connection active, any network traffic between the two hosts is encrypted. To configure a host-to-host IPsec connection, use the following steps for each host:

Note You should perform the following procedures on the actual machine that you are configuring. Avoid attempting to configure and establish IPsec connections remotely.

100

IPsec Host-to-Host Configuration

1. In a command shell, type system-config-network to start the Network Administration Tool. 2. On the IPsec tab, click New to start the IPsec configuration wizard. 3. Click Forward to start configuring a host-to-host IPsec connection. 4. Enter a unique name for the connection, for example, ipsec0. If required, select the check box to automatically activate the connection when the computer starts. Click Forward to continue. 5. Select Host to Host encryption as the connection type, and then click Forward. 6. Select the type of encryption to use: manual or automatic. If you select manual encryption, an encryption key must be provided later in the process. If you select automatic encryption, the racoon daemon manages the encryption key. The ipsectools package must be installed if you want to use automatic encryption. Click Forward to continue. 7. Enter the IP address of the remote host. To determine the IP address of the remote host, use the following command on the remote host: [root@myServer ~] # /sbin/ifconfig

where is the Ethernet device that you want to use for the VPN connection. If only one Ethernet card exists in the system, the device name is typically eth0. The following example shows the relevant information from this command (note that this is an example output only): eth0

Link encap:Ethernet HWaddr 00:0C:6E:E8:98:1D inet addr:172.16.44.192 Bcast:172.16.45.255 Mask:255.255.254.0

The IP address is the number following the inet addr: label.

Note For host-to-host connections, both hosts should have a public, routable address. Alternatively, both hosts can have a private, non-routable address (for example, from the 10.x.x.x or 192.168.x.x ranges) as long as they are on the same LAN. If the hosts are on different LANs, or one has a public address while the other has a private address, refer to Section 2.7.7, “IPsec Network-to-Network Configuration”. Click Forward to continue. 8. If manual encryption was selected in step 6, specify the encryption key to use, or click Generate to create one. a. Specify an authentication key or click Generate to generate one. It can be any combination of numbers and letters.

101


b. Click Forward to continue. 9. Verify the information on the IPsec — Summary page, and then click Apply. 10. Click File => Save to save the configuration. You may need to restart the network for the changes to take effect. To restart the network, use the following command: [root@myServer ~]# service network restart

11. Select the IPsec connection from the list and click the Activate button. 12. Repeat the entire procedure for the other host. It is essential that the same keys from step 8 be used on the other hosts. Otherwise, IPsec will not work. After configuring the IPsec connection, it appears in the IPsec list as shown in Figure 2.10, “IPsec Connection”.

Figure 2.10. IPsec Connection The following files are created when the IPsec connection is configured: • /etc/sysconfig/network-scripts/ifcfg- • /etc/sysconfig/network-scripts/keys- • /etc/racoon/.conf

102


• /etc/racoon/psk.txt If automatic encryption is selected, /etc/racoon/racoon.conf is also created. When the interface is up, /etc/racoon/racoon.conf is modified to include .conf.

2.7.6.2. Manual IPsec Host-to-Host Configuration The first step in creating a connection is to gather system and network information from each workstation. For a host-to-host connection, you need the following: • The IP address of each host • A unique name, for example, ipsec1. This is used to identify the IPsec connection and to distinguish it from other devices or connections. • A fixed encryption key or one automatically generated by racoon. • A pre-shared authentication key that is used during the initial stage of the connection and to exchange encryption keys during the session. For example, suppose Workstation A and Workstation B want to connect to each other through an IPsec tunnel. They want to connect using a pre-shared key with the value of Key_Value01, and the users agree to let racoon automatically generate and share an authentication key between each host. Both host users decide to name their connections ipsec1.

Note You should choose a PSK that uses a mixture of upper- and lower-case characters, numbers and punctuation. An easily-guessable PSK constitutes a security risk. It is not necessary to use the same connection name for each host. You should choose a name that is convenient and meaningful for your installation. The following is the IPsec configuration file for Workstation A for a host-to-host IPsec connection with Workstation B. The unique name to identify the connection in this example is ipsec1, so the resulting file is called /etc/sysconfig/network-scripts/ifcfg-ipsec1. DST=X.X.X.XTYPE=IPSEC ONBOOT=no IKE_METHOD=PSK

For Workstation A, X.X.X.X is the IP address of Workstation B. For Workstation B, X.X.X.X is the IP address of Workstation A. This connection is not set to initiate on boot-up (ONBOOT=no) and it uses the pre-shared key method of authentication (IKE_METHOD=PSK). The following is the content of the pre-shared key file (called /etc/sysconfig/networkscripts/keys-ipsec1) that both workstations need to authenticate each other. The contents of this file should be identical on both workstations, and only the root user should be able to read or write this file. IKE_PSK=Key_Value01

103


Important To change the keys-ipsec1 file so that only the root user can read or edit the file, use the following command after creating the file: [root@myServer ~] # chmod 600 /etc/sysconfig/network-scripts/keys-ipsec1

To change the authentication key at any time, edit the keys-ipsec1 file on both workstations. Both authentication keys must be identical for proper connectivity. The next example shows the specific configuration for the phase 1 connection to the remote host. The file is called X.X.X.X.conf, where X.X.X.X is the IP address of the remote IPsec host. Note that this file is automatically generated when the IPsec tunnel is activated and should not be edited directly. remote X.X.X.X{ exchange_mode aggressive, main; my_identifier address; proposal { encryption_algorithm 3des; hash_algorithm sha1; authentication_method pre_shared_key; dh_group 2 ; } }

The default phase 1 configuration file that is created when an IPsec connection is initialized contains the following statements used by the Fedora implementation of IPsec: remote X.X.X.X Specifies that the subsequent stanzas of this configuration file apply only to the remote node identified by the X.X.X.X IP address. exchange_mode aggressive The default configuration for IPsec on Fedora uses an aggressive authentication mode, which lowers the connection overhead while allowing configuration of several IPsec connections with multiple hosts. my_identifier address Specifies the identification method to use when authenticating nodes. Fedora uses IP addresses to identify nodes. encryption_algorithm 3des Specifies the encryption cipher used during authentication. By default, Triple Data Encryption Standard (3DES) is used. hash_algorithm sha1; Specifies the hash algorithm used during phase 1 negotiation between nodes. By default, Secure Hash Algorithm version 1 is used. authentication_method pre_shared_key Specifies the authentication method used during node negotiation. By default, Fedora uses preshared keys for authentication.

104


dh_group 2 Specifies the Diffie-Hellman group number for establishing dynamically-generated session keys. By default, modp1024 (group 2) is used.

2.7.6.2.1. The Racoon Configuration File The /etc/racoon/racoon.conf files should be identical on all IPsec nodes except for the include "/etc/racoon/X.X.X.X.conf" statement. This statement (and the file it references) is generated when the IPsec tunnel is activated. For Workstation A, the X.X.X.X in the include statement is Workstation B's IP address. The opposite is true of Workstation B. The following shows a typical racoon.conf file when the IPsec connection is activated. # Racoon IKE daemon configuration file. # See 'man racoon.conf' for a description of the format and entries. path include "/etc/racoon"; path pre_shared_key "/etc/racoon/psk.txt"; path certificate "/etc/racoon/certs"; sainfo anonymous { pfs_group 2; lifetime time 1 hour ; encryption_algorithm 3des, blowfish 448, rijndael ; authentication_algorithm hmac_sha1, hmac_md5 ; compression_algorithm deflate ; } include "/etc/racoon/X.X.X.X.conf";

This default racoon.conf file includes defined paths for IPsec configuration, pre-shared key files, and certificates. The fields in sainfo anonymous describe the phase 2 SA between the IPsec nodes — the nature of the IPsec connection (including the supported encryption algorithms used) and the method of exchanging keys. The following list defines the fields of phase 2: sainfo anonymous Denotes that SA can anonymously initialize with any peer provided that the IPsec credentials match. pfs_group 2 Defines the Diffie-Hellman key exchange protocol, which determines the method by which the IPsec nodes establish a mutual temporary session key for the second phase of IPsec connectivity. By default, the Fedora implementation of IPsec uses group 2 (or modp1024) of the DiffieHellman cryptographic key exchange groups. Group 2 uses a 1024-bit modular exponentiation that prevents attackers from decrypting previous IPsec transmissions even if a private key is compromised. lifetime time 1 hour This parameter specifies the lifetime of an SA and can be quantified either by time or by bytes of data. The default Fedora implementation of IPsec specifies a one hour lifetime. encryption_algorithm 3des, blowfish 448, rijndael Specifies the supported encryption ciphers for phase 2. Fedora supports 3DES, 448-bit Blowfish, and Rijndael (the cipher used in the Advanced Encryption Standard, or AES).

105


authentication_algorithm hmac_sha1, hmac_md5 Lists the supported hash algorithms for authentication. Supported modes are sha1 and md5 hashed message authentication codes (HMAC). compression_algorithm deflate Defines the Deflate compression algorithm for IP Payload Compression (IPCOMP) support, which allows for potentially faster transmission of IP datagrams over slow connections. To start the connection, use the following command on each host: [root@myServer ~]# /sbin/ifup

where is the name you specified for the IPsec connection. To test the IPsec connection, run the tcpdump utility to view the network packets being transfered between the hosts and verify that they are encrypted via IPsec. The packet should include an AH header and should be shown as ESP packets. ESP means it is encrypted. For example: [root@myServer ~]# tcpdump -n -i eth0 host IP 172.16.45.107 > 172.16.44.192: AH(spi=0x0954ccb6,seq=0xbb): ESP(spi=0x0c9f2164,seq=0xbb)

2.7.7. IPsec Network-to-Network Configuration IPsec can also be configured to connect an entire network (such as a LAN or WAN) to a remote network using a network-to-network connection. A network-to-network connection requires the setup of IPsec routers on each side of the connecting networks to transparently process and route information from one node on a LAN to a node on a remote LAN. Figure 2.11, “A network-to-network IPsec tunneled connection” shows a network-to-network IPsec tunneled connection.

Figure 2.11. A network-to-network IPsec tunneled connection This diagram shows two separate LANs separated by the Internet. These LANs use IPsec routers to authenticate and initiate a connection using a secure tunnel through the Internet. Packets that are intercepted in transit would require brute-force decryption in order to crack the cipher protecting the packets between these LANs. The process of communicating from one node in the 192.168.1.0/24 IP range to another in the 192.168.2.0/24 range is completely transparent to the nodes as the processing, encryption/decryption, and routing of the IPsec packets are completely handled by the IPsec router. The information needed for a network-to-network connection include: • The externally-accessible IP addresses of the dedicated IPsec routers

106

IPsec Network-to-Network Configuration

• The network address ranges of the LAN/WAN served by the IPsec routers (such as 192.168.1.0/24 or 10.0.1.0/24) • The IP addresses of the gateway devices that route the data from the network nodes to the Internet • A unique name, for example, ipsec1. This is used to identify the IPsec connection and to distinguish it from other devices or connections. • A fixed encryption key or one automatically generated by racoon • A pre-shared authentication key that is used during the initial stage of the connection and to exchange encryption keys during the session.

2.7.7.1. Network-to-Network (VPN) Connection A network-to-network IPsec connection uses two IPsec routers, one for each network, through which the network traffic for the private subnets is routed. For example, as shown in Figure 2.12, “Network-to-Network IPsec”, if the 192.168.1.0/24 private network sends network traffic to the 192.168.2.0/24 private network, the packets go through gateway0, to ipsec0, through the Internet, to ipsec1, to gateway1, and to the 192.168.2.0/24 subnet. IPsec routers require publicly addressable IP addresses and a second Ethernet device connected to their respective private networks. Traffic only travels through an IPsec router if it is intended for another IPsec router with which it has an encrypted connection.

Figure 2.12. Network-to-Network IPsec Alternate network configuration options include a firewall between each IP router and the Internet, and an intranet firewall between each IPsec router and subnet gateway. The IPsec router and the gateway for the subnet can be one system with two Ethernet devices: one with a public IP address that acts as the IPsec router; and one with a private IP address that acts as the gateway for the private subnet. Each IPsec router can use the gateway for its private network or a public gateway to send the packets to the other IPsec router. Use the following procedure to configure a network-to-network IPsec connection: 1. In a command shell, type system-config-network to start the Network Administration Tool. 2. On the IPsec tab, click New to start the IPsec configuration wizard. 3. Click Forward to start configuring a network-to-network IPsec connection. 4. Enter a unique nickname for the connection, for example, ipsec0. If required, select the check box to automatically activate the connection when the computer starts. Click Forward to continue.

107


5. Select Network to Network encryption (VPN) as the connection type, and then click Forward. 6. Select the type of encryption to use: manual or automatic. If you select manual encryption, an encryption key must be provided later in the process. If you select automatic encryption, the racoon daemon manages the encryption key. The ipsectools package must be installed if you want to use automatic encryption. Click Forward to continue. 7. On the Local Network page, enter the following information: • Local Network Address — The IP address of the device on the IPsec router connected to the private network. • Local Subnet Mask — The subnet mask of the local network IP address. • Local Network Gateway — The gateway for the private subnet. Click Forward to continue.

Figure 2.13. Local Network Information 8. On the Remote Network page, enter the following information: • Remote IP Address — The publicly addressable IP address of the IPsec router for the other private network. In our example, for ipsec0, enter the publicly addressable IP address of ipsec1, and vice versa.

108


• Remote Network Address — The network address of the private subnet behind the other IPsec router. In our example, enter 192.168.1.0 if configuring ipsec1, and enter 192.168.2.0 if configuring ipsec0. • Remote Subnet Mask — The subnet mask of the remote IP address. • Remote Network Gateway — The IP address of the gateway for the remote network address. • If manual encryption was selected in step 6, specify the encryption key to use or click Generate to create one. Specify an authentication key or click Generate to generate one. This key can be any combination of numbers and letters. Click Forward to continue.

Figure 2.14. Remote Network Information 9. Verify the information on the IPsec — Summary page, and then click Apply. 10. Select File => Save to save the configuration. 11. Select the IPsec connection from the list, and then click Activate to activate the connection. 12. Enable IP forwarding: a. Edit /etc/sysctl.conf and set net.ipv4.ip_forward to 1.

109


b. Use the following command to enable the change: [root@myServer ~]# /sbin/sysctl -p /etc/sysctl.conf

The network script to activate the IPsec connection automatically creates network routes to send packets through the IPsec router if necessary.

2.7.7.2. Manual IPsec Network-to-Network Configuration Suppose LAN A (lana.example.com) and LAN B (lanb.example.com) want to connect to each other through an IPsec tunnel. The network address for LAN A is in the 192.168.1.0/24 range, while LAN B uses the 192.168.2.0/24 range. The gateway IP address is 192.168.1.254 for LAN A and 192.168.2.254 for LAN B. The IPsec routers are separate from each LAN gateway and use two network devices: eth0 is assigned to an externally-accessible static IP address which accesses the Internet, while eth1 acts as a routing point to process and transmit LAN packets from one network node to the remote network nodes. The IPsec connection between each network uses a pre-shared key with the value of r3dh4tl1nux, and the administrators of A and B agree to let racoon automatically generate and share an authentication key between each IPsec router. The administrator of LAN A decides to name the IPsec connection ipsec0, while the administrator of LAN B names the IPsec connection ipsec1. The following example shows the contents of the ifcfg file for a network-to-network IPsec connection for LAN A. The unique name to identify the connection in this example is ipsec0, so the resulting file is called /etc/sysconfig/network-scripts/ifcfg-ipsec0. TYPE=IPSEC ONBOOT=yes IKE_METHOD=PSK SRCGW=192.168.1.254 DSTGW=192.168.2.254 SRCNET=192.168.1.0/24 DSTNET=192.168.2.0/24 DST=X.X.X.X

The following list describes the contents of this file: TYPE=IPSEC Specifies the type of connection. ONBOOT=yes Specifies that the connection should initiate on boot-up. IKE_METHOD=PSK Specifies that the connection uses the pre-shared key method of authentication. SRCGW=192.168.1.254 The IP address of the source gateway. For LAN A, this is the LAN A gateway, and for LAN B, the LAN B gateway. DSTGW=192.168.2.254 The IP address of the destination gateway. For LAN A, this is the LAN B gateway, and for LAN B, the LAN A gateway.

110


SRCNET=192.168.1.0/24 Specifies the source network for the IPsec connection, which in this example is the network range for LAN A. DSTNET=192.168.2.0/24 Specifies the destination network for the IPsec connection, which in this example is the network range for LAN B. DST=X.X.X.X The externally-accessible IP address of LAN B. The following example is the content of the pre-shared key file called /etc/sysconfig/networkscripts/keys-ipsecX (where X is 0 for LAN A and 1 for LAN B) that both networks use to authenticate each other. The contents of this file should be identical and only the root user should be able to read or write this file. IKE_PSK=r3dh4tl1nux

Important To change the keys-ipsecX file so that only the root user can read or edit the file, use the following command after creating the file: chmod 600 /etc/sysconfig/network-scripts/keys-ipsec1

To change the authentication key at any time, edit the keys-ipsecX file on both IPsec routers. Both keys must be identical for proper connectivity. The following example is the contents of the /etc/racoon/racoon.conf configuration file for the IPsec connection. Note that the include line at the bottom of the file is automatically generated and only appears if the IPsec tunnel is running. # Racoon IKE daemon configuration file. # See 'man racoon.conf' for a description of the format and entries. path include "/etc/racoon"; path pre_shared_key "/etc/racoon/psk.txt"; path certificate "/etc/racoon/certs"; sainfo anonymous { pfs_group 2; lifetime time 1 hour ; encryption_algorithm 3des, blowfish 448, rijndael ; authentication_algorithm hmac_sha1, hmac_md5 ; compression_algorithm deflate ; } include "/etc/racoon/X.X.X.X.conf"

The following is the specific configuration for the connection to the remote network. The file is called X.X.X.X.conf (where X.X.X.X is the IP address of the remote IPsec router). Note that this file is automatically generated when the IPsec tunnel is activated and should not be edited directly. remote X.X.X.X{

111


exchange_mode aggressive, main; my_identifier address; proposal { encryption_algorithm 3des; hash_algorithm sha1; authentication_method pre_shared_key; dh_group 2 ; } }

Prior to starting the IPsec connection, IP forwarding should be enabled in the kernel. To enable IP forwarding: 1. Edit /etc/sysctl.conf and set net.ipv4.ip_forward to 1. 2. Use the following command to enable the change: [root@myServer ~] # sysctl -p /etc/sysctl.conf

To start the IPsec connection, use the following command on each router: [root@myServer ~] # /sbin/ifup ipsec0

The connections are activated, and both LAN A and LAN B are able to communicate with each other. The routes are created automatically via the initialization script called by running ifup on the IPsec connection. To show a list of routes for the network, use the following command: [root@myServer ~] # /sbin/ip route list

To test the IPsec connection, run the tcpdump utility on the externally-routable device (eth0 in this example) to view the network packets being transfered between the hosts (or networks), and verify that they are encrypted via IPsec. For example, to check the IPsec connectivity of LAN A, use the following command: [root@myServer ~] # tcpdump -n -i eth0 host lana.example.com

The packet should include an AH header and should be shown as ESP packets. ESP means it is encrypted. For example (back slashes denote a continuation of one line): 12:24:26.155529 lanb.example.com > lana.example.com: AH(spi=0x021c9834,seq=0x358): \ lanb.example.com > lana.example.com: ESP(spi=0x00c887ad,seq=0x358) (DF) \ (ipip-proto-4)

2.7.8. Starting and Stopping an IPsec Connection If the IPsec connection was not configured to activate on boot, you can control it from the command line. To start the connection, use the following command on each host for host-to-host IPsec, or each IPsec router for network-to-network IPsec: [root@myServer ~] # /sbin/ifup

where is the nickname configured earlier, such as ipsec0.

112

Firewalls

To stop the connection, use the following command: [root@myServer ~] # /sbin/ifdown

2.8. Firewalls Information security is commonly thought of as a process and not a product. However, standard security implementations usually employ some form of dedicated mechanism to control access privileges and restrict network resources to users who are authorized, identifiable, and traceable. Fedora includes several tools to assist administrators and security engineers with network-level access control issues. Firewalls are one of the core components of a network security implementation. Several vendors market firewall solutions catering to all levels of the marketplace: from home users protecting one PC to data center solutions safeguarding vital enterprise information. Firewalls can be stand-alone hardware solutions, such as firewall appliances by Cisco, Nokia, and Sonicwall. Vendors such as Checkpoint, McAfee, and Symantec have also developed proprietary software firewall solutions for home and business markets. Apart from the differences between hardware and software firewalls, there are also differences in the way firewalls function that separate one solution from another. Table 2.2, “Firewall Types” details three common types of firewalls and how they function: Method Description

Advantages

Disadvantages

NAT

Network Address Translation (NAT) places private IP subnetworks behind one or a small pool of public IP addresses, masquerading all requests to one source rather than several. The Linux kernel has built-in NAT functionality through the Netfilter kernel subsystem.

· Can be configured transparently to machines on a LAN · Protection of many machines and services behind one or more external IP addresses simplifies administration duties · Restriction of user access to and from the LAN can be configured by opening and closing ports on the NAT firewall/gateway

· Cannot prevent malicious activity once users connect to a service outside of the firewall

Packet Filter

A packet filtering firewall reads each data packet that passes through a LAN. It can read and process packets by header information and filters the packet based on sets of programmable rules implemented by the firewall administrator. The Linux kernel has built-in packet filtering functionality through the Netfilter kernel subsystem.

· Customizable through the iptables front-end utility · Does not require any customization on the client side, as all network activity is filtered at the router level rather than the application level · Since packets are not transmitted through a proxy, network performance is faster due to direct connection from client to remote host

· Cannot filter packets for content like proxy firewalls · Processes packets at the protocol layer, but cannot filter packets at an application layer · Complex network architectures can make establishing packet filtering rules difficult, especially if coupled with IP masquerading or local subnets and DMZ networks

113


Method Description

Advantages

Disadvantages

Proxy

· Gives administrators control over what applications and protocols function outside of the LAN · Some proxy servers can cache frequently-accessed data locally rather than having to use the Internet connection to request it. This helps to reduce bandwidth consumption · Proxy services can be logged and monitored closely, allowing tighter control over resource utilization on the network

· Proxies are often application-specific (HTTP, Telnet, etc.), or protocolrestricted (most proxies work with TCP-connected services only) · Application services cannot run behind a proxy, so your application servers must use a separate form of network security · Proxies can become a network bottleneck, as all requests and transmissions are passed through one source rather than directly from a client to a remote service

Proxy firewalls filter all requests of a certain protocol or type from LAN clients to a proxy machine, which then makes those requests to the Internet on behalf of the local client. A proxy machine acts as a buffer between malicious remote users and the internal network client machines.

Table 2.2. Firewall Types

2.8.1. Netfilter and IPTables The Linux kernel features a powerful networking subsystem called Netfilter. The Netfilter subsystem provides stateful or stateless packet filtering as well as NAT and IP masquerading services. Netfilter also has the ability to mangle IP header information for advanced routing and connection state management. Netfilter is controlled using the iptables tool.

2.8.1.1. IPTables Overview The power and flexibility of Netfilter is implemented using the iptables administration tool, a command line tool similar in syntax to its predecessor, ipchains, which Netfilter/iptables replaced in the Linux kernel 2.4 and above. iptables uses the Netfilter subsystem to enhance network connection, inspection, and processing. iptables features advanced logging, pre- and post-routing actions, network address translation, and port forwarding, all in one command line interface. This section provides an overview of iptables. For more detailed information, refer to Section 2.9, “IPTables”.

2.8.2. Basic Firewall Configuration Just as a firewall in a building attempts to prevent a fire from spreading, a computer firewall attempts to prevent malicious software from spreading to your computer. It also helps to prevent unauthorized users from accessing your computer. In a default Fedora installation, a firewall exists between your computer or network and any untrusted networks, for example the Internet. It determines which services on your computer remote users can access. A properly configured firewall can greatly increase the security of your system. It is recommended that you configure a firewall for any Fedora system with an Internet connection.

114

Basic Firewall Configuration

2.8.2.1. Firewall Configuration Tool During the Firewall Configuration screen of the Fedora installation, you were given the option to enable a basic firewall as well as to allow specific devices, incoming services, and ports. After installation, you can change this preference by using the Firewall Configuration Tool. To start this application, use the following command: [root@myServer ~] # system-config-firewall

Figure 2.15. Firewall Configuration Tool

Note The Firewall Configuration Tool only configures a basic firewall. If the system needs more complex rules, refer to Section 2.9, “IPTables” for details on configuring specific iptables rules.

2.8.2.2. Enabling and Disabling the Firewall Select one of the following options for the firewall: • Disabled — Disabling the firewall provides complete access to your system and does no security checking. This should only be selected if you are running on a trusted network (not the Internet) or need to configure a custom firewall using the iptables command line tool.

115


Warning Firewall configurations and any customized firewall rules are stored in the / etc/sysconfig/iptables file. If you choose Disabled and click OK, these configurations and firewall rules will be lost. • Enabled — This option configures the system to reject incoming connections that are not in response to outbound requests, such as DNS replies or DHCP requests. If access to services running on this machine is needed, you can choose to allow specific services through the firewall. If you are connecting your system to the Internet, but do not plan to run a server, this is the safest choice.

2.8.2.3. Trusted Services Enabling options in the Trusted services list allows the specified service to pass through the firewall. WWW (HTTP) The HTTP protocol is used by Apache (and by other Web servers) to serve web pages. If you plan on making your Web server publicly available, select this check box. This option is not required for viewing pages locally or for developing web pages. This service requires that the httpd package be installed. Enabling WWW (HTTP) will not open a port for HTTPS, the SSL version of HTTP. If this service is required, select the Secure WWW (HTTPS) check box. FTP The FTP protocol is used to transfer files between machines on a network. If you plan on making your FTP server publicly available, select this check box. This service requires that the vsftpd package be installed. SSH Secure Shell (SSH) is a suite of tools for logging into and executing commands on a remote machine. To allow remote access to the machine via ssh, select this check box. This service requires that the openssh-server package be installed. Telnet Telnet is a protocol for logging into remote machines. Telnet communications are unencrypted and provide no security from network snooping. Allowing incoming Telnet access is not recommended. To allow remote access to the machine via telnet, select this check box. This service requires that the telnet-server package be installed. Mail (SMTP) SMTP is a protocol that allows remote hosts to connect directly to your machine to deliver mail. You do not need to enable this service if you collect your mail from your ISP's server using POP3 or IMAP, or if you use a tool such as fetchmail. To allow delivery of mail to your machine, select this check box. Note that an improperly configured SMTP server can allow remote machines to use your server to send spam.

116

Using IPTables

NFS4 The Network File System (NFS) is a file sharing protocol commonly used on *NIX systems. Version 4 of this protocol is more secure than its predecessors. If you want to share files or directories on your system with other network users, select this check box. Samba Samba is an implementation of Microsoft's proprietary SMB networking protocol. If you need to share files, directories, or locally-connected printers with Microsoft Windows machines, select this check box.

2.8.2.4. Other Ports The Firewall Configuration Tool includes an Other ports section for specifying custom IP ports as being trusted by iptables. For example, to allow IRC and Internet printing protocol (IPP) to pass through the firewall, add the following to the Other ports section: 194:tcp,631:tcp

2.8.2.5. Saving the Settings Click OK to save the changes and enable or disable the firewall. If Enable firewall was selected, the options selected are translated to iptables commands and written to the /etc/sysconfig/ iptables file. The iptables service is also started so that the firewall is activated immediately after saving the selected options. If Disable firewall was selected, the /etc/sysconfig/iptables file is removed and the iptables service is stopped immediately. The selected options are also written to the /etc/sysconfig/system-config-securitylevel file so that the settings can be restored the next time the application is started. Do not edit this file by hand. Even though the firewall is activated immediately, the iptables service is not configured to start automatically at boot time. Refer to Section 2.8.2.6, “Activating the IPTables Service” for more information.

2.8.2.6. Activating the IPTables Service The firewall rules are only active if the iptables service is running. To manually start the service, use the following command: [root@myServer ~] # service iptables restart

To ensure that iptables starts when the system is booted, use the following command: [root@myServer ~] # chkconfig --level 345 iptables on

2.8.3. Using IPTables The first step in using iptables is to start the iptables service. Use the following command to start the iptables service: [root@myServer ~] # service iptables start

117


Note The ip6tables service can be turned off if you intend to use the iptables service only. If you deactivate the ip6tables service, remember to deactivate the IPv6 network also. Never leave a network device active without the matching firewall. To force iptables to start by default when the system is booted, use the following command: [root@myServer ~] # chkconfig --level 345 iptables on

This forces iptables to start whenever the system is booted into runlevel 3, 4, or 5.

2.8.3.1. IPTables Command Syntax The following sample iptables command illustrates the basic command syntax: [root@myServer ~ ] # iptables -A -j

The -A option specifies that the rule be appended to . Each chain is comprised of one or more rules, and is therefore also known as a ruleset. The three built-in chains are INPUT, OUTPUT, and FORWARD. These chains are permanent and cannot be deleted. The chain specifies the point at which a packet is manipulated. The -j option specifies the target of the rule; i.e., what to do if the packet matches the rule. Examples of built-in targets are ACCEPT, DROP, and REJECT. Refer to the iptables man page for more information on the available chains, options, and targets.

2.8.3.2. Basic Firewall Policies Establishing basic firewall policies creates a foundation for building more detailed, user-defined rules. Each iptables chain is comprised of a default policy, and zero or more rules which work in concert with the default policy to define the overall ruleset for the firewall. The default policy for a chain can be either DROP or ACCEPT. Security-minded administrators typically implement a default policy of DROP, and only allow specific packets on a case-by-case basis. For example, the following policies block all incoming and outgoing packets on a network gateway: [root@myServer ~ ] # iptables -P INPUT DROP [root@myServer ~ ] # iptables -P OUTPUT DROP

It is also recommended that any forwarded packets — network traffic that is to be routed from the firewall to its destination node — be denied as well, to restrict internal clients from inadvertent exposure to the Internet. To do this, use the following rule: [root@myServer ~ ] # iptables -P FORWARD DROP

When you have established the default policies for each chain, you can create and save further rules for your particular network and security requirements. The following sections describe how to save iptables rules and outline some of the rules you might implement in the course of building your iptables firewall.

118

Common IPTables Filtering

2.8.3.3. Saving and Restoring IPTables Rules Changes to iptables are transitory; if the system is rebooted or if the iptables service is restarted, the rules are automatically flushed and reset. To save the rules so that they are loaded when the iptables service is started, use the following command: [root@myServer ~ ] # service iptables save

The rules are stored in the file /etc/sysconfig/iptables and are applied whenever the service is started or the machine is rebooted.

2.8.4. Common IPTables Filtering Preventing remote attackers from accessing a LAN is one of the most important aspects of network security. The integrity of a LAN should be protected from malicious remote users through the use of stringent firewall rules. However, with a default policy set to block all incoming, outgoing, and forwarded packets, it is impossible for the firewall/gateway and internal LAN users to communicate with each other or with external resources. To allow users to perform network-related functions and to use networking applications, administrators must open certain ports for communication. For example, to allow access to port 80 on the firewall, append the following rule: [root@myServer ~ ] # iptables -A INPUT -p tcp -m tcp --dport 80 -j ACCEPT

This allows users to browse websites that communicate using the standard port 80. To allow access to secure websites (for example, https://www.example.com/), you also need to provide access to port 443, as follows: [root@myServer ~ ] # iptables -A INPUT -p tcp -m tcp --dport 443 -j ACCEPT

Important When creating an iptables ruleset, order is important. If a rule specifies that any packets from the 192.168.100.0/24 subnet be dropped, and this is followed by a rule that allows packets from 192.168.100.13 (which is within the dropped subnet), then the second rule is ignored. The rule to allow packets from 192.168.100.13 must precede the rule that drops the remainder of the subnet. To insert a rule in a specific location in an existing chain, use the -I option. For example: [root@myServer ~ ] # iptables -I INPUT 1 -i lo -p all -j ACCEPT

This rule is inserted as the first rule in the INPUT chain to allow local loopback device traffic.

119


There may be times when you require remote access to the LAN. Secure services, for example SSH, can be used for encrypted remote connection to LAN services. Administrators with PPP-based resources (such as modem banks or bulk ISP accounts), dial-up access can be used to securely circumvent firewall barriers. Because they are direct connections, modem connections are typically behind a firewall/gateway. For remote users with broadband connections, however, special cases can be made. You can configure iptables to accept connections from remote SSH clients. For example, the following rules allow remote SSH access: [root@myServer ~ ] # iptables -A INPUT -p tcp --dport 22 -j ACCEPT [root@myServer ~ ] # iptables -A OUTPUT -p tcp --sport 22 -j ACCEPT

These rules allow incoming and outbound access for an individual system, such as a single PC directly connected to the Internet or a firewall/gateway. However, they do not allow nodes behind the firewall/gateway to access these services. To allow LAN access to these services, you can use Network Address Translation (NAT) with iptables filtering rules.

2.8.5. FORWARD and NAT Rules Most ISPs provide only a limited number of publicly routable IP addresses to the organizations they serve. Administrators must, therefore, find alternative ways to share access to Internet services without giving public IP addresses to every node on the LAN. Using private IP addresses is the most common way of allowing all nodes on a LAN to properly access internal and external network services. Edge routers (such as firewalls) can receive incoming transmissions from the Internet and route the packets to the intended LAN node. At the same time, firewalls/gateways can also route outgoing requests from a LAN node to the remote Internet service. This forwarding of network traffic can become dangerous at times, especially with the availability of modern cracking tools that can spoof internal IP addresses and make the remote attacker's machine act as a node on your LAN. To prevent this, iptables provides routing and forwarding policies that can be implemented to prevent abnormal usage of network resources. The FORWARD chain allows an administrator to control where packets can be routed within a LAN. For example, to allow forwarding for the entire LAN (assuming the firewall/gateway is assigned an internal IP address on eth1), use the following rules: [root@myServer ~ ] # iptables -A FORWARD -i eth1 -j ACCEPT [root@myServer ~ ] # iptables -A FORWARD -o eth1 -j ACCEPT

This rule gives systems behind the firewall/gateway access to the internal network. The gateway routes packets from one LAN node to its intended destination node, passing all packets through its eth1 device.

120

FORWARD and NAT Rules

Note By default, the IPv4 policy in Fedora kernels disables support for IP forwarding. This prevents machines that run Fedora from functioning as dedicated edge routers. To enable IP forwarding, use the following command: [root@myServer ~ ] # sysctl -w net.ipv4.ip_forward=1

This configuration change is only valid for the current session; it does not persist beyond a reboot or network service restart. To permanently set IP forwarding, edit the /etc/sysctl.conf file as follows: Locate the following line: net.ipv4.ip_forward = 0

Edit it to read as follows: net.ipv4.ip_forward = 1

Use the following command to enable the change to the sysctl.conf file: [root@myServer ~ ] # sysctl -p /etc/sysctl.conf

2.8.5.1. Postrouting and IP Masquerading Accepting forwarded packets via the firewall's internal IP device allows LAN nodes to communicate with each other; however they still cannot communicate externally to the Internet. To allow LAN nodes with private IP addresses to communicate with external public networks, configure the firewall for IP masquerading, which masks requests from LAN nodes with the IP address of the firewall's external device (in this case, eth0): [root@myServer ~ ] # iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE

This rule uses the NAT packet matching table (-t nat) and specifies the built-in POSTROUTING chain for NAT (-A POSTROUTING) on the firewall's external networking device (-o eth0). POSTROUTING allows packets to be altered as they are leaving the firewall's external device. The -j MASQUERADE target is specified to mask the private IP address of a node with the external IP address of the firewall/gateway.

2.8.5.2. Prerouting If you have a server on your internal network that you want make available externally, you can use the -j DNAT target of the PREROUTING chain in NAT to specify a destination IP address and port where incoming packets requesting a connection to your internal service can be forwarded. For example, if you want to forward incoming HTTP requests to your dedicated Apache HTTP Server at 172.31.0.23, use the following command:

121


[root@myServer ~ ] # iptables -t nat -A PREROUTING -i eth0 -p tcp --dport 80 -j DNAT --to 172.31.0.23:80

This rule specifies that the nat table use the built-in PREROUTING chain to forward incoming HTTP requests exclusively to the listed destination IP address of 172.31.0.23.

Note If you have a default policy of DROP in your FORWARD chain, you must append a rule to forward all incoming HTTP requests so that destination NAT routing is possible. To do this, use the following command: [root@myServer ~ ] # iptables -A FORWARD -i eth0 -p tcp --dport 80 -d 172.31.0.23 -j ACCEPT

This rule forwards all incoming HTTP requests from the firewall to the intended destination; the Apache HTTP Server behind the firewall.

2.8.5.3. DMZs and IPTables You can create iptables rules to route traffic to certain machines, such as a dedicated HTTP or FTP server, in a demilitarized zone (DMZ). A DMZ is a special local subnetwork dedicated to providing services on a public carrier, such as the Internet. For example, to set a rule for routing incoming HTTP requests to a dedicated HTTP server at 10.0.4.2 (outside of the 192.168.1.0/24 range of the LAN), NAT uses the PREROUTING table to forward the packets to the appropriate destination: [root@myServer ~ ] # iptables -t nat -A PREROUTING -i eth0 -p tcp --dport 80 -j DNAT --todestination 10.0.4.2:80

With this command, all HTTP connections to port 80 from outside of the LAN are routed to the HTTP server on a network separate from the rest of the internal network. This form of network segmentation can prove safer than allowing HTTP connections to a machine on the network. If the HTTP server is configured to accept secure connections, then port 443 must be forwarded as well.

2.8.6. Malicious Software and Spoofed IP Addresses More elaborate rules can be created that control access to specific subnets, or even specific nodes, within a LAN. You can also restrict certain dubious applications or programs such as trojans, worms, and other client/server viruses from contacting their server. For example, some trojans scan networks for services on ports from 31337 to 31340 (called the elite ports in cracking terminology). Since there are no legitimate services that communicate via these non-standard ports, blocking them can effectively diminish the chances that potentially infected nodes on your network independently communicate with their remote master servers. The following rules drop all TCP traffic that attempts to use port 31337:

122

IPTables and Connection Tracking

[root@myServer ~ ] # iptables -A OUTPUT -o eth0 -p tcp --dport 31337 --sport 31337 -j DROP [root@myServer ~ ] # iptables -A FORWARD -o eth0 -p tcp --dport 31337 --sport 31337 -j DROP

You can also block outside connections that attempt to spoof private IP address ranges to infiltrate your LAN. For example, if your LAN uses the 192.168.1.0/24 range, you can design a rule that instructs the Internet-facing network device (for example, eth0) to drop any packets to that device with an address in your LAN IP range. Because it is recommended to reject forwarded packets as a default policy, any other spoofed IP address to the external-facing device (eth0) is rejected automatically. [root@myServer ~ ] # iptables -A FORWARD -s 192.168.1.0/24 -i eth0 -j DROP

Note There is a distinction between the DROP and REJECT targets when dealing with appended rules. The REJECT target denies access and returns a connection refused error to users who attempt to connect to the service. The DROP target, as the name implies, drops the packet without any warning. Administrators can use their own discretion when using these targets. However, to avoid user confusion and attempts to continue connecting, the REJECT target is recommended.

2.8.7. IPTables and Connection Tracking You can inspect and restrict connections to services based on their connection state. A module within iptables uses a method called connection tracking to store information about incoming connections. You can allow or deny access based on the following connection states: • NEW — A packet requesting a new connection, such as an HTTP request. • ESTABLISHED — A packet that is part of an existing connection. • RELATED — A packet that is requesting a new connection but is part of an existing connection. For example, FTP uses port 21 to establish a connection, but data is transferred on a different port (typically port 20). • INVALID — A packet that is not part of any connections in the connection tracking table. You can use the stateful functionality of iptables connection tracking with any network protocol, even if the protocol itself is stateless (such as UDP). The following example shows a rule that uses connection tracking to forward only the packets that are associated with an established connection: [root@myServer ~ ] # iptables -A FORWARD -m state --state ESTABLISHED,RELATED -j ACCEPT

123


2.8.8. IPv6 The introduction of the next-generation Internet Protocol, called IPv6, expands beyond the 32-bit address limit of IPv4 (or IP). IPv6 supports 128-bit addresses, and carrier networks that are IPv6 aware are therefore able to address a larger number of routable addresses than IPv4. Fedora supports IPv6 firewall rules using the Netfilter 6 subsystem and the ip6tables command. In Fedora 12, both IPv4 and IPv6 services are enabled by default. The ip6tables command syntax is identical to iptables in every aspect except that it supports 128-bit addresses. For example, use the following command to enable SSH connections on an IPv6aware network server: [root@myServer ~ ] # ip6tables -A INPUT -i eth0 -p tcp -s 3ffe:ffff:100::1/128 --dport 22 -j ACCEPT

For more information about IPv6 networking, refer to the IPv6 Information Page at http:// www.ipv6.org/.

2.8.9. Additional Resources There are several aspects to firewalls and the Linux Netfilter subsystem that could not be covered in this chapter. For more information, refer to the following resources.

2.8.9.1. Installed Firewall Documentation • Refer to Section 2.9, “IPTables” for more detailed information on the iptables command, including definitions for many command options. • The iptables man page contains a brief summary of the various options.

2.8.9.2. Useful Firewall Websites • http://www.netfilter.org/ — The official homepage of the Netfilter and iptables project. • http://www.tldp.org/ — The Linux Documentation Project contains several useful guides relating to firewall creation and administration. • http://www.iana.org/assignments/port-numbers — The official list of registered and common service ports as assigned by the Internet Assigned Numbers Authority.

2.8.9.3. Related Documentation • Red Hat Linux Firewalls, by Bill McCarty; Red Hat Press — a comprehensive reference to building network and server firewalls using open source packet filtering technology such as Netfilter and iptables. It includes topics that cover analyzing firewall logs, developing firewall rules, and customizing your firewall using various graphical tools. • Linux Firewalls, by Robert Ziegler; New Riders Press — contains a wealth of information on building firewalls using both 2.2 kernel ipchains as well as Netfilter and iptables. Additional security topics such as remote access issues and intrusion detection systems are also covered.

124

IPTables

2.9. IPTables Included with Fedora are advanced tools for network packet filtering — the process of controlling network packets as they enter, move through, and exit the network stack within the kernel. Kernel versions prior to 2.4 relied on ipchains for packet filtering and used lists of rules applied to packets at each step of the filtering process. The 2.4 kernel introduced iptables (also called netfilter), which is similar to ipchains but greatly expands the scope and control available for filtering network packets. This chapter focuses on packet filtering basics, explains various options available with iptables commands, and explains how filtering rules can be preserved between system reboots. Refer to Section 2.9.6, “Additional Resources” for instructions on how to construct iptables rules and setting up a firewall based on these rules.

Important The default firewall mechanism in the 2.4 and later kernels is iptables, but iptables cannot be used if ipchains is already running. If ipchains is present at boot time, the kernel issues an error and fails to start iptables. The functionality of ipchains is not affected by these errors.

2.9.1. Packet Filtering The Linux kernel uses the Netfilter facility to filter packets, allowing some of them to be received by or pass through the system while stopping others. This facility is built in to the Linux kernel, and has three built-in tables or rules lists, as follows: • filter — The default table for handling network packets. • nat — Used to alter packets that create a new connection and used for Network Address Translation (NAT). • mangle — Used for specific types of packet alteration. Each table has a group of built-in chains, which correspond to the actions performed on the packet by netfilter. The built-in chains for the filter table are as follows: • INPUT — Applies to network packets that are targeted for the host. • OUTPUT — Applies to locally-generated network packets. • FORWARD — Applies to network packets routed through the host. The built-in chains for the nat table are as follows: • PREROUTING — Alters network packets when they arrive. • OUTPUT — Alters locally-generated network packets before they are sent out. • POSTROUTING — Alters network packets before they are sent out.

125


The built-in chains for the mangle table are as follows: • INPUT — Alters network packets targeted for the host. • OUTPUT — Alters locally-generated network packets before they are sent out. • FORWARD — Alters network packets routed through the host. • PREROUTING — Alters incoming network packets before they are routed. • POSTROUTING — Alters network packets before they are sent out. Every network packet received by or sent from a Linux system is subject to at least one table. However, a packet may be subjected to multiple rules within each table before emerging at the end of the chain. The structure and purpose of these rules may vary, but they usually seek to identify a packet coming from or going to a particular IP address, or set of addresses, when using a particular protocol and network service.

Note By default, firewall rules are saved in the /etc/sysconfig/iptables or /etc/ sysconfig/ip6tables files. The iptables service starts before any DNS-related services when a Linux system is booted. This means that firewall rules can only reference numeric IP addresses (for example, 192.168.0.1). Domain names (for example, host.example.com) in such rules produce errors. Regardless of their destination, when packets match a particular rule in one of the tables, a target or action is applied to them. If the rule specifies an ACCEPT target for a matching packet, the packet skips the rest of the rule checks and is allowed to continue to its destination. If a rule specifies a DROP target, that packet is refused access to the system and nothing is sent back to the host that sent the packet. If a rule specifies a QUEUE target, the packet is passed to user-space. If a rule specifies the optional REJECT target, the packet is dropped, but an error packet is sent to the packet's originator. Every chain has a default policy to ACCEPT, DROP, REJECT, or QUEUE. If none of the rules in the chain apply to the packet, then the packet is dealt with in accordance with the default policy. The iptables command configures these tables, as well as sets up new tables if necessary.

2.9.2. Command Options for IPTables Rules for filtering packets are created using the iptables command. The following aspects of the packet are most often used as criteria: • Packet Type — Specifies the type of packets the command filters. • Packet Source/Destination — Specifies which packets the command filters based on the source or destination of the packet. • Target — Specifies what action is taken on packets matching the above criteria. Refer to Section 2.9.2.4, “IPTables Match Options” and Section 2.9.2.5, “Target Options” for more information about specific options that address these aspects of a packet.

126

Command Options for IPTables

The options used with specific iptables rules must be grouped logically, based on the purpose and conditions of the overall rule, for the rule to be valid. The remainder of this section explains commonlyused options for the iptables command.

2.9.2.1. Structure of IPTables Command Options Many iptables commands have the following structure: iptables [-t ] \ \

— Specifies which table the rule applies to. If omitted, the filter table is used. — Specifies the action to perform, such as appending or deleting a rule. — Specifies the chain to edit, create, or delete. - pairs — Parameters and associated options that specify how to process a packet that matches the rule. The length and complexity of an iptables command can change significantly, based on its purpose. For example, a command to remove a rule from a chain can be very short: iptables -D In contrast, a command that adds a rule which filters packets from a particular subnet using a variety of specific parameters and options can be rather long. When constructing iptables commands, it is important to remember that some parameters and options require further parameters and options to construct a valid rule. This can produce a cascading effect, with the further parameters requiring yet more parameters. Until every parameter and option that requires another set of options is satisfied, the rule is not valid. Type iptables -h to view a comprehensive list of iptables command structures.

2.9.2.2. Command Options Command options instruct iptables to perform a specific action. Only one command option is allowed per iptables command. With the exception of the help command, all commands are written in upper-case characters. The iptables commands are as follows: • -A — Appends the rule to the end of the specified chain. Unlike the -I option described below, it does not take an integer argument. It always appends the rule to the end of the specified chain. • -C — Checks a particular rule before adding it to the user-specified chain. This command can help you construct complex iptables rules by prompting you for additional parameters and options. • -D | — Deletes a rule in a particular chain by number (such as 5 for the fifth rule in a chain), or by rule specification. The rule specification must exactly match an existing rule. • -E — Renames a user-defined chain. A user-defined chain is any chain other than the default, preexisting chains. (Refer to the -N option, below, for information on creating user-defined chains.) This is a cosmetic change and does not affect the structure of the table.

127


Note If you attempt to rename one of the default chains, the system reports a Match not found error. You cannot rename the default chains. • -F — Flushes the selected chain, which effectively deletes every rule in the chain. If no chain is specified, this command flushes every rule from every chain. • -h — Provides a list of command structures, as well as a quick summary of command parameters and options. • -I [] — Inserts the rule in the specified chain at a point specified by a user-defined integer argument. If no argument is specified, the rule is inserted at the top of the chain.

Important As noted above, the order of rules in a chain determines which rules apply to which packets. This is important to remember when adding rules using either the -A or -I option. This is especially important when adding rules using the -I with an integer argument. If you specify an existing number when adding a rule to a chain, iptables adds the new rule before (or above) the existing rule. • -L — Lists all of the rules in the chain specified after the command. To list all rules in all chains in the default filter table, do not specify a chain or table. Otherwise, the following syntax should be used to list the rules in a specific chain in a particular table: iptables -L -t

Additional options for the -L command option, which provide rule numbers and allow more verbose rule descriptions, are described in Section 2.9.2.6, “Listing Options”. • -N — Creates a new chain with a user-specified name. The chain name must be unique, otherwise an error message is displayed. • -P — Sets the default policy for the specified chain, so that when packets traverse an entire chain without matching a rule, they are sent to the specified target, such as ACCEPT or DROP. • -R — Replaces a rule in the specified chain. The rule's number must be specified after the chain's name. The first rule in a chain corresponds to rule number one. • -X — Deletes a user-specified chain. You cannot delete a built-in chain. • -Z — Sets the byte and packet counters in all chains for a table to zero.

2.9.2.3. IPTables Parameter Options Certain iptables commands, including those used to add, append, delete, insert, or replace rules within a particular chain, require various parameters to construct a packet filtering rule.

128


• -c — Resets the counters for a particular rule. This parameter accepts the PKTS and BYTES options to specify which counter to reset. • -d — Sets the destination hostname, IP address, or network of a packet that matches the rule. When matching a network, the following IP address/netmask formats are supported: • N.N.N.N/M.M.M.M — Where N.N.N.N is the IP address range and M.M.M.M is the netmask. • N.N.N.N/M — Where N.N.N.N is the IP address range and M is the bitmask. • -f — Applies this rule only to fragmented packets. You can use the exclamation point character (!) option after this parameter to specify that only unfragmented packets are matched.

Note Distinguishing between fragmented and unfragmented packets is desirable, despite fragmented packets being a standard part of the IP protocol. Originally designed to allow IP packets to travel over networks with differing frame sizes, these days fragmentation is more commonly used to generate DoS attacks using mal-formed packets. It's also worth noting that IPv6 disallows fragmentation entirely. • -i — Sets the incoming network interface, such as eth0 or ppp0. With iptables, this optional parameter may only be used with the INPUT and FORWARD chains when used with the filter table and the PREROUTING chain with the nat and mangle tables. This parameter also supports the following special options: • Exclamation point character (!) — Reverses the directive, meaning any specified interfaces are excluded from this rule. • Plus character (+) — A wildcard character used to match all interfaces that match the specified string. For example, the parameter -i eth+ would apply this rule to any Ethernet interfaces but exclude any other interfaces, such as ppp0. If the -i parameter is used but no interface is specified, then every interface is affected by the rule. • -j — Jumps to the specified target when a packet matches a particular rule. The standard targets are ACCEPT, DROP, QUEUE, and RETURN. Extended options are also available through modules loaded by default with the Fedora iptables RPM package. Valid targets in these modules include LOG, MARK, and REJECT, among others. Refer to the iptables man page for more information about these and other targets. This option can also be used to direct a packet matching a particular rule to a user-defined chain outside of the current chain so that other rules can be applied to the packet. If no target is specified, the packet moves past the rule with no action taken. The counter for this rule, however, increases by one.

129


• -o — Sets the outgoing network interface for a rule. This option is only valid for the OUTPUT and FORWARD chains in the filter table, and the POSTROUTING chain in the nat and mangle tables. This parameter accepts the same options as the incoming network interface parameter (-i). • -p — Sets the IP protocol affected by the rule. This can be either icmp, tcp, udp, or all, or it can be a numeric value, representing one of these or a different protocol. You can also use any protocols listed in the /etc/protocols file. The "all" protocol means the rule applies to every supported protocol. If no protocol is listed with this rule, it defaults to "all". • -s — Sets the source for a particular packet using the same syntax as the destination (-d) parameter.

2.9.2.4. IPTables Match Options Different network protocols provide specialized matching options which can be configured to match a particular packet using that protocol. However, the protocol must first be specified in the iptables command. For example, -p enables options for the specified protocol. Note that you can also use the protocol ID, instead of the protocol name. Refer to the following examples, each of which have the same effect: iptables -A INPUT -p icmp --icmp-type any -j ACCEPT

iptables -A INPUT -p 5813 --icmp-type any -j ACCEPT

Service definitions are provided in the /etc/services file. For readability, it is recommended that you use the service names rather than the port numbers.

Warning Secure the /etc/services file to prevent unauthorized editing. If this file is editable, crackers can use it to enable ports on your machine you have otherwise closed. To secure this file, type the following commands as root:

[root@myServer ~]# chown root.root /etc/services [root@myServer ~]# chmod 0644 /etc/services [root@myServer ~]# chattr +i /etc/services

This prevents the file from being renamed, deleted or having links made to it.

2.9.2.4.1. TCP Protocol These match options are available for the TCP protocol (-p tcp): • --dport — Sets the destination port for the packet. To configure this option, use a network service name (such as www or smtp); a port number; or a range of port numbers. To specify a range of port numbers, separate the two numbers with a colon (:). For example: -p tcp --dport 3000:3200. The largest acceptable valid range is 0:65535.

130


Use an exclamation point character (!) after the --dport option to match all packets that do not use that network service or port. To browse the names and aliases of network services and the port numbers they use, view the / etc/services file. The --destination-port match option is synonymous with --dport. • --sport — Sets the source port of the packet using the same options as --dport. The -source-port match option is synonymous with --sport. • --syn — Applies to all TCP packets designed to initiate communication, commonly called SYN packets. Any packets that carry a data payload are not touched. Use an exclamation point character (!) after the --syn option to match all non-SYN packets. • --tcp-flags — Allows TCP packets that have specific bits (flags) set, to match a rule. The --tcp-flags match option accepts two parameters. The first parameter is the mask; a comma-separated list of flags to be examined in the packet. The second parameter is a commaseparated list of flags that must be set for the rule to match. The possible flags are: • ACK • FIN • PSH • RST • SYN • URG • ALL • NONE For example, an iptables rule that contains the following specification only matches TCP packets that have the SYN flag set and the ACK and FIN flags not set: --tcp-flags ACK,FIN,SYN SYN Use the exclamation point character (!) after the --tcp-flags to reverse the effect of the match option. • --tcp-option — Attempts to match with TCP-specific options that can be set within a particular packet. This match option can also be reversed with the exclamation point character (!).

2.9.2.4.2. UDP Protocol These match options are available for the UDP protocol (-p udp):

131


• --dport — Specifies the destination port of the UDP packet, using the service name, port number, or range of port numbers. The --destination-port match option is synonymous with --dport. • --sport — Specifies the source port of the UDP packet, using the service name, port number, or range of port numbers. The --source-port match option is synonymous with --sport. For the --dport and --sport options, to specify a range of port numbers, separate the two numbers with a colon (:). For example: -p tcp --dport 3000:3200. The largest acceptable valid range is 0:65535.

2.9.2.4.3. ICMP Protocol The following match options are available for the Internet Control Message Protocol (ICMP) (-p icmp): • --icmp-type — Sets the name or number of the ICMP type to match with the rule. A list of valid ICMP names can be retrieved by typing the iptables -p icmp -h command.

2.9.2.4.4. Additional Match Option Modules Additional match options are available through modules loaded by the iptables command. To use a match option module, load the module by name using the -m , where is the name of the module. Many modules are available by default. You can also create modules to provide additional functionality. The following is a partial list of the most commonly used modules: • limit module — Places limits on how many packets are matched to a particular rule. When used in conjunction with the LOG target, the limit module can prevent a flood of matching packets from filling up the system log with repetitive messages or using up system resources. Refer to Section 2.9.2.5, “Target Options” for more information about the LOG target. The limit module enables the following options: • --limit — Sets the maximum number of matches for a particular time period, specified as a / pair. For example, using --limit 5/hour allows five rule matches per hour. Periods can be specified in seconds, minutes, hours, or days. If a number and time modifier are not used, the default value of 3/hour is assumed. • --limit-burst — Sets a limit on the number of packets able to match a rule at one time. This option is specified as an integer and should be used in conjunction with the --limit option. If no value is specified, the default value of five (5) is assumed. • state module — Enables state matching. The state module enables the following options:

132


• --state — match a packet with the following connection states: • ESTABLISHED — The matching packet is associated with other packets in an established connection. You need to accept this state if you want to maintain a connection between a client and a server. • INVALID — The matching packet cannot be tied to a known connection. • NEW — The matching packet is either creating a new connection or is part of a two-way connection not previously seen. You need to accept this state if you want to allow new connections to a service. • RELATED — The matching packet is starting a new connection related in some way to an existing connection. An example of this is FTP, which uses one connection for control traffic (port 21), and a separate connection for data transfer (port 20). These connection states can be used in combination with one another by separating them with commas, such as -m state --state INVALID,NEW. • mac module — Enables hardware MAC address matching. The mac module enables the following option: • --mac-source — Matches a MAC address of the network interface card that sent the packet. To exclude a MAC address from a rule, place an exclamation point character (!) after the --macsource match option. Refer to the iptables man page for more match options available through modules.

2.9.2.5. Target Options When a packet has matched a particular rule, the rule can direct the packet to a number of different targets which determine the appropriate action. Each chain has a default target, which is used if none of the rules on that chain match a packet or if none of the rules which match the packet specify a target. The following are the standard targets: • — A user-defined chain within the table. User-defined chain names must be unique. This target passes the packet to the specified chain. • ACCEPT — Allows the packet through to its destination or to another chain. • DROP — Drops the packet without responding to the requester. The system that sent the packet is not notified of the failure. • QUEUE — The packet is queued for handling by a user-space application. • RETURN — Stops checking the packet against rules in the current chain. If the packet with a RETURN target matches a rule in a chain called from another chain, the packet is returned to the first chain to resume rule checking where it left off. If the RETURN rule is used on a built-in chain and the packet cannot move up to its previous chain, the default target for the current chain is used. In addition, extensions are available which allow other targets to be specified. These extensions are called target modules or match option modules and most only apply to specific tables and situations.

133


Refer to Section 2.9.2.4.4, “Additional Match Option Modules” for more information about match option modules. Many extended target modules exist, most of which only apply to specific tables or situations. Some of the most popular target modules included by default in Fedora are: • LOG — Logs all packets that match this rule. Because the packets are logged by the kernel, the / etc/syslog.conf file determines where these log entries are written. By default, they are placed in the /var/log/messages file. Additional options can be used after the LOG target to specify the way in which logging occurs: • --log-level — Sets the priority level of a logging event. Refer to the syslog.conf man page for a list of priority levels. • --log-ip-options — Logs any options set in the header of an IP packet. • --log-prefix — Places a string of up to 29 characters before the log line when it is written. This is useful for writing syslog filters for use in conjunction with packet logging.

Note Due to an issue with this option, you should add a trailing space to the logprefix value. • --log-tcp-options — Logs any options set in the header of a TCP packet. • --log-tcp-sequence — Writes the TCP sequence number for the packet in the log. • REJECT — Sends an error packet back to the remote system and drops the packet. The REJECT target accepts --reject-with (where is the rejection type) allowing more detailed information to be returned with the error packet. The message portunreachable is the default error type given if no other option is used. Refer to the iptables man page for a full list of options. Other target extensions, including several that are useful for IP masquerading using the nat table, or with packet alteration using the mangle table, can be found in the iptables man page.

2.9.2.6. Listing Options The default list command, iptables -L [], provides a very basic overview of the default filter table's current chains. Additional options provide more information: • -v — Displays verbose output, such as the number of packets and bytes each chain has processed, the number of packets and bytes each rule has matched, and which interfaces apply to a particular rule. • -x — Expands numbers into their exact values. On a busy system, the number of packets and bytes processed by a particular chain or rule may be abbreviated to Kilobytes, Megabytes (Megabytes) or Gigabytes. This option forces the full number to be displayed. • -n — Displays IP addresses and port numbers in numeric format, rather than the default hostname and network service format.

134

Saving IPTables Rules

• --line-numbers — Lists rules in each chain next to their numeric order in the chain. This option is useful when attempting to delete the specific rule in a chain or to locate where to insert a rule within a chain. • -t — Specifies a table name. If omitted, defaults to the filter table.

2.9.3. Saving IPTables Rules Rules created with the iptables command are stored in memory. If the system is restarted before saving the iptables rule set, all rules are lost. For netfilter rules to persist through a system reboot, they need to be saved. To save netfilter rules, type the following command as root: /sbin/service iptables save

This executes the iptables init script, which runs the /sbin/iptables-save program and writes the current iptables configuration to /etc/sysconfig/iptables. The existing /etc/ sysconfig/iptables file is saved as /etc/sysconfig/iptables.save. The next time the system boots, the iptables init script reapplies the rules saved in /etc/ sysconfig/iptables by using the /sbin/iptables-restore command. While it is always a good idea to test a new iptables rule before committing it to the /etc/ sysconfig/iptables file, it is possible to copy iptables rules into this file from another system's version of this file. This provides a quick way to distribute sets of iptables rules to multiple machines. You can also save the iptables rules to a separate file for distribution, backup or other purposes. To save your iptables rules, type the following command as root: [root@myServer ~]# iptables-save > where is a user-defined name for your ruleset.

Important If distributing the /etc/sysconfig/iptables file to other machines, type /sbin/ service iptables restart for the new rules to take effect.

Note Note the difference between the iptables command (/sbin/iptables), which is used to manipulate the tables and chains that constitute the iptables functionality, and the iptables service (/sbin/iptables service), which is used to enable and disable the iptables service itself.

2.9.4. IPTables Control Scripts There are two basic methods for controlling iptables in Fedora: • Firewall Configuration Tool (system-config-securitylevel) — A graphical interface for creating, activating, and saving basic firewall rules. Refer to Section 2.8.2, “Basic Firewall Configuration” for more information.

135


• /sbin/service iptables — Used to manipulate various functions of iptables using its initscript. The following options are available: • start — If a firewall is configured (that is, /etc/sysconfig/iptables exists), all running iptables are stopped completely and then started using the /sbin/iptables-restore command. This option only works if the ipchains kernel module is not loaded. To check if this module is loaded, type the following command as root: [root@MyServer ~]# lsmod | grep ipchains

If this command returns no output, it means the module is not loaded. If necessary, use the / sbin/rmmod command to remove the module. • stop — If a firewall is running, the firewall rules in memory are flushed, and all iptables modules and helpers are unloaded. If the IPTABLES_SAVE_ON_STOP directive in the /etc/sysconfig/iptables-config configuration file is changed from its default value to yes, current rules are saved to /etc/ sysconfig/iptables and any existing rules are moved to the file /etc/sysconfig/ iptables.save. Refer to Section 2.9.4.1, “IPTables Control Scripts Configuration File” for more information about the iptables-config file. • restart — If a firewall is running, the firewall rules in memory are flushed, and the firewall is started again if it is configured in /etc/sysconfig/iptables. This option only works if the ipchains kernel module is not loaded. If the IPTABLES_SAVE_ON_RESTART directive in the /etc/sysconfig/iptables-config configuration file is changed from its default value to yes, current rules are saved to /etc/ sysconfig/iptables and any existing rules are moved to the file /etc/sysconfig/ iptables.save. Refer to Section 2.9.4.1, “IPTables Control Scripts Configuration File” for more information about the iptables-config file. • status — Displays the status of the firewall and lists all active rules. The default configuration for this option displays IP addresses in each rule. To display domain and hostname information, edit the /etc/sysconfig/iptables-config file and change the value of IPTABLES_STATUS_NUMERIC to no. Refer to Section 2.9.4.1, “IPTables Control Scripts Configuration File” for more information about the iptables-config file. • panic — Flushes all firewall rules. The policy of all configured tables is set to DROP. This option could be useful if a server is known to be compromised. Rather than physically disconnecting from the network or shutting down the system, you can use this option to stop all further network traffic but leave the machine in a state ready for analysis or other forensics. • save — Saves firewall rules to /etc/sysconfig/iptables using iptables-save. Refer to Section 2.9.3, “Saving IPTables Rules” for more information.

136

IPTables Control Scripts

Note To use the same initscript commands to control netfilter for IPv6, substitute ip6tables for iptables in the /sbin/service commands listed in this section. For more information about IPv6 and netfilter, refer to Section 2.9.5, “IPTables and IPv6”.

2.9.4.1. IPTables Control Scripts Configuration File The behavior of the iptables initscripts is controlled by the /etc/sysconfig/iptables-config configuration file. The following is a list of directives contained in this file: • IPTABLES_MODULES — Specifies a space-separated list of additional iptables modules to load when a firewall is activated. These can include connection tracking and NAT helpers. • IPTABLES_MODULES_UNLOAD — Unloads modules on restart and stop. This directive accepts the following values: • yes — The default value. This option must be set to achieve a correct state for a firewall restart or stop. • no — This option should only be set if there are problems unloading the netfilter modules. • IPTABLES_SAVE_ON_STOP — Saves current firewall rules to /etc/sysconfig/iptables when the firewall is stopped. This directive accepts the following values: • yes — Saves existing rules to /etc/sysconfig/iptables when the firewall is stopped, moving the previous version to the /etc/sysconfig/iptables.save file. • no — The default value. Does not save existing rules when the firewall is stopped. • IPTABLES_SAVE_ON_RESTART — Saves current firewall rules when the firewall is restarted. This directive accepts the following values: • yes — Saves existing rules to /etc/sysconfig/iptables when the firewall is restarted, moving the previous version to the /etc/sysconfig/iptables.save file. • no — The default value. Does not save existing rules when the firewall is restarted. • IPTABLES_SAVE_COUNTER — Saves and restores all packet and byte counters in all chains and rules. This directive accepts the following values: • yes — Saves the counter values. • no — The default value. Does not save the counter values. • IPTABLES_STATUS_NUMERIC — Outputs IP addresses in numeric form instead of domain or hostnames. This directive accepts the following values: • yes — The default value. Returns only IP addresses within a status output. • no — Returns domain or hostnames within a status output.

137


2.9.5. IPTables and IPv6 If the iptables-ipv6 package is installed, netfilter in Fedora can filter the next-generation IPv6 Internet protocol. The command used to manipulate the IPv6 netfilter is ip6tables. Most directives for this command are identical to those used for iptables, except the nat table is not yet supported. This means that it is not yet possible to perform IPv6 network address translation tasks, such as masquerading and port forwarding. Rules for ip6tables are saved in the /etc/sysconfig/ip6tables file. Previous rules saved by the ip6tables initscripts are saved in the /etc/sysconfig/ip6tables.save file. Configuration options for the ip6tables init script are stored in /etc/sysconfig/ip6tablesconfig, and the names for each directive vary slightly from their iptables counterparts. For example, the iptables-config directive IPTABLES_MODULES:the equivalent in the ip6tables-config file is IP6TABLES_MODULES.

2.9.6. Additional Resources Refer to the following sources for additional information on packet filtering with iptables. • Section 2.8, “Firewalls” — Contains a chapter about the role of firewalls within an overall security strategy as well as strategies for constructing firewall rules.

2.9.6.1. Installed IP Tables Documentation • man iptables — Contains a description of iptables as well as a comprehensive list of targets, options, and match extensions.

2.9.6.2. Useful IP Tables Websites • http://www.netfilter.org/ — The home of the netfilter/iptables project. Contains assorted information about iptables, including a FAQ addressing specific problems and various helpful guides by Rusty Russell, the Linux IP firewall maintainer. The HOWTO documents on the site cover subjects such as basic networking concepts, kernel packet filtering, and NAT configurations. • http://www.linuxnewbie.org/nhf/Security/IPtables_Basics.html — An introduction to the way packets move through the Linux kernel, plus an introduction to constructing basic iptables commands.

138

Chapter 3.

Encryption There are two main types of data that must be protected: data at rest and data in motion. These different types of data are protected in similar ways using similar technology but the implementations can be completely different. No single protective implementation can prevent all possible methods of compromise as the same information may be at rest and in motion at different points in time.

3.1. Data at Rest Data at rest is data that is stored on a hard drive, tape, CD, DVD, disk, or other media. This information's biggest threat comes from being physically stolen. Laptops in airports, CDs going through the mail, and backup tapes that get left in the wrong places are all examples of events where data can be compromised through theft. If the data was encrypted on the media then you wouldn't have to worry as much about the data being compromised.

3.2. Full Disk Encryption Full disk or partition encryption is one of the best ways of protecting your data. Not only is each file protected but also the temporary storage that may contain parts of these files is also protected. Full disk encryption will protect all of your files so you don't have to worry about selecting what you want to protect and possibly missing a file. Fedora 9, and later, natively supports LUKS Encryption. LUKS will bulk encrypt your hard drive partitions so that while your computer is off your data is protected. This will also protect your computer from attackers attempting to use single-user-mode to login to your computer or otherwise gain access. Full disk encryption solutions like LUKS only protect the data when your computer is off. Once the computer is on and LUKS has decrypted the disk, the files on that disk are available to anyone who would normally have access to them. To protect your files when the computer is on, use full disk encryption in combination with another solution such as file based encryption. Also remember to lock your computer whenever you are away from it. A passphrase protected screen saver set to activate after a few minutes of inactivity is a good way to keep intruders out.

3.3. File Based Encryption GnuPG (GPG) is an open source version of PGP that allows you to sign and/or encrypt a file or an email message. This is useful to maintain integrity of the message or file and also protects the confidentiality of the information contained within the file or email. In the case of email, GPG provides dual protection. Not only can it provide Data at Rest protection but also Data In Motion protection once the message has been sent across the network. File based encryption is intended to protect a file after it has left your computer, such as when you send a CD through the mail. Some file based encryption solutions will leave remnants of the encrypted files that an attacker who has physical access to your computer can recover under some circumstances. To protect the contents of those files from attackers who may have access to your computer, use file based encryption combined with another solution such as full disk encryption.

3.4. Data in Motion Data in motion is data that is being transmitted over a network. The biggest threats to data in motion are interception and alteration. Your user name and password should never be transmitted over a network without protection as it could be intercepted and used by someone else to impersonate you

139

Chapter 3. Encryption

or gain access to sensitive information. Other private information such as bank account information should also be protected when transmitted across a network. If the network session was encrypted then you would not have to worry as much about the data being compromised while it is being transmitted. Data in motion is particularly vulnerable to attackers because the attacker does not have to be near the computer in which the data is being stored rather they only have to be somewhere along the path. Encryption tunnels can protect data along the path of communications.

3.5. Virtual Private Networks Virtual Private Networks (VPN) provide encrypted tunnels between computers or networks of computers across all ports. With a VPN in place, all network traffic from the client is forwarded to the server through the encrypted tunnel. This means that the client is logically on the same network as the server it is connected to via the VPN. VPNs are very common and are simple to use and setup.

3.6. Secure Shell Secure Shell (SSH) is a powerful network protocol used to communicate with another system over a secure channel. The transmissions over SSH are encrypted and protected from interception. Cryptographic log-on can also be utilized to provide a better authentication method over traditional usernames and passwords. SSH is very easy to activate. By simply starting the sshd service, the system will begin to accept connections and will allow access to the system when a correct username and password is provided during the connection process. The standard TCP port for the SSH service is 22, however this can be changed by modifying the configuration file /etc/ssh/sshd_config and restarting the service. This file also contains other configuration options for SSH. Secure Shell (SSH) also provides encrypted tunnels between computers but only using a single port. 1 Port forwarding can be done over an SSH tunnel and traffic will be encrypted as it passes over that tunnel but using port forwarding is not as fluid as a VPN.

3.7. LUKS Disk Encryption Linux Unified Key Setup-on-disk-format (or LUKS) allows you to encrypt partitions on your Linux computer. This is particularly important when it comes to mobile computers and removable media. LUKS allows multiple user keys to decrypt a master key which is used for the bulk encryption of the partition.

3.7.1. LUKS Implementation in Fedora Fedora 9, and later, utilizes LUKS to perform file system encryption. By default, the option to encrypt the file system is unchecked during the installation. If you select the option to encrypt you hard drive, you will be prompted for a passphrase that will be asked every time you boot the computer. This passphrase "unlocks" the bulk encryption key that is used to decrypt your partition. If you choose to modify the default partition table you can choose which partitions you want to encrypt. This is set in the partition table settings Fedora's default implementation of LUKS is AES 128 with a SHA256 hashing. Ciphers that are available are: 1

http://www.redhatmagazine.com/2007/11/27/advanced-ssh-configuration-and-tunneling-we-dont-need-no-stinking-vpn-software

140

Manually Encrypting Directories

2

• AES - Advanced Encryption Standard - FIPS PUB 197 • Twofish (A 128-bit Block Cipher) • Serpent 3

• cast5 - RFC 2144

4

• cast6 - RFC 2612

3.7.2. Manually Encrypting Directories Warning Following this procedure will remove all data on the partition that you are encrypting. You WILL lose all your information! Make sure you backup your data to an external source before beginning this procedure! If you are running a version of Fedora prior to Fedora 9 and want to encrypt a partition, or you want to encrypt a partition after the installation of the current version of Fedora, the following directions are for you. The below example demonstrates encrypting your /home partition but any partition can be used. The following procedure will wipe all your existing data, so be sure to have a tested backup before you start. This also requires you to have a separate partition for /home (in my case that is /dev/VG00/ LV_home). All the following must be done as root. Any of these steps failing means you must not continue until the step succeeded.

3.7.3. Step-by-Step Instructions 1. enter runlevel 1: telinit 1 2. unmount your existing /home: umount /home 3. if it fails use fuser to find and kill processes hogging /home: fuser -mvk /home 4. verify /home is not mounted any longer: cat /proc/mounts | grep home 5. Fill your partition with random data: dd if=/dev/urandom of=/dev/VG00/LV_home This process takes many hours to complete.

Important The process, however, is imperative in order to have good protection against break-in attempts. Just let it run overnight. 6. initialize your partition: cryptsetup --verbose --verify-passphrase luksFormat / dev/VG00/LV_home 7. open the newly encrypted device: cryptsetup luksOpen /dev/VG00/LV_home home 8. check it's there: ls -l /dev/mapper | grep home

141


9. create a filesystem: mkfs.ext3 /dev/mapper/home 10. mount it: mount /dev/mapper/home /home 11. check it's visible: df -h | grep home 12. add the following to /etc/crypttab: home /dev/VG00/LV_home none 13. edit your /etc/fstab, removing the old entry for /home and adding /dev/mapper/home /home ext3 defaults 1 2 14. verify your fstab entry: mount /home 15. restore default SELinux security contexts: /sbin/restorecon -v -R /home 16. reboot: shutdown -r now 17. The entry in /etc/crypttab makes your computer ask your luks passphrase on boot 18. Login as root and restore your backup

3.7.4. What you have just accomplished. Congratulations, you now have an encrypted partition for all of your data to safely rest while the computer is off.

3.7.5. Links of Interest For additional information on LUKS or encrypting hard drives under Fedora please visit one of the following links: 5

• LUKS - Linux Unified Key Setup

• HOWTO: Creating an encrypted Physical Volume (PV) using a second hard drive, pvmove, and a 6 Fedora LiveCD

3.8. 7-Zip Encrypted Archives 7

7-Zip is a cross-platform, next generation, file compression tool that can also use strong encryption (AES-256) to protect the contents of the archive. This is extremely useful when you need to move data between multiple computers that use varying operating systems (i.e. Linux at home, Windows at work) and you want a portable encryption solution.

3.8.1. 7-Zip Installation in Fedora 7-Zip is not a base package in Fedora, but it is available in the software repository. Once installed, the package will update alongside the rest of the software on the computer with no special attention necessary.

3.8.2. Step-by-Step Installation Instructions • Open a Terminal: Click ''Applications'' -> ''System Tools'' -> ''Terminal'' 7

http://www.7-zip.org/

142

Step-by-Step Usage Instructions

• Install 7-Zip with sudo access: sudo yum install p7zip • Close the Terminal: exit

3.8.3. Step-by-Step Usage Instructions By following these instructions you are going to compress and encrypt your "Documents" directory. Your original "Documents" directory will remain unaltered. This technique can be applied to any directory or file you have access to on the filesystem. • Open a Terminal:Click ''Applications'' -> ''System Tools'' -> ''Terminal'' • Compress and Encrypt: (enter a password when prompted) 7za a -mhe=on -ms=on -p Documents.7z Documents/ The "Documents" directory is now compressed and encrypted. The following instructions will move the encrypted archive somewhere new and then extract it. • Create a new directory: mkdir newplace • Move the encrypted file: mv Documents.7z newplace • Go to the new directory: cd newplace • Extract the file: (enter the password when prompted) 7za x Documents.7z The archive is now extracted into the new location. The following instructions will clean up all the prior steps and restore your computer to its previous state. • Go up a directory: cd .. • Delete the test archive and test extraction: rm -r newplace • Close the Terminal: exit

3.8.4. Things of note 7-Zip is not shipped by default with Microsoft Windows or Mac OS X. If you need to use your 7-Zip files on those platforms you will need to install the appropriate version of 7-Zip on those computers. 8 See the 7-Zip download page .

3.9. Using GNU Privacy Guard (GnuPG) GnuPG (GPG) is used to identify yourself and authenticate your communications, including those with people you don't know. GPG allows anyone reading a GPG-signed email to verify its authenticity. In other words, GPG allows someone to be reasonably certain that communications signed by you actually are from you. GPG is useful because it helps prevent third parties from altering code or intercepting conversations and altering the message. GPG can also be used to sign and/or encrypt files kept on your computer or on a network drive. This can add additional protection in preventing a file from being altered or read by unauthorized people. To utilize GPG for authentication or encrytion of email you must first generate your public and private keys. After generating the keys you will have to setup your email client to utilize them. 8

http://www.7-zip.org/download.html

143


3.9.1. Generating GPG Keys in GNOME Install the Seahorse utility, which makes GPG key management easier. From the main menu, select System > Administration > Add/Remove Software and wait for PackageKit to start. Enter Seahorse into the text box and select the Find. Select the checkbox next to the ''seahorse'' package and select ''Apply'' to add the software. You can also install Seahorse at the command line with the command su -c "yum install seahorse". To create a key, from the ''Applications > Accessories'' menu select ''Passwords and Encryption Keys'', which starts the application Seahorse. From the ''Key'' menu select ''Create New Key...'' then ''PGP Key'' then click ''Continue''. Type your full name, email address, and an optional comment describing who are you (e.g.: John C. Smith, [email protected], The Man). Click ''Create''. A dialog is displayed asking for a passphrase for the key. Choose a strong passphrase but also easy to remember. Click ''OK'' and the key is created.

Warning If you forget your passphrase, the key cannot be used and any data encrypted using that key will be lost. To find your GPG key ID, look in the ''Key ID'' column next to the newly created key. In most cases, if you are asked for the key ID, you should prepend "0x" to the key ID, as in "0x6789ABCD". You should make a backup of your private key and store it somewhere secure.

3.9.2. Generating GPG Keys in KDE Start the KGpg program from the main menu by selecting Applications > Utilities > Encryption Tool. If you have never used KGpg before, the program walks you through the process of creating your own GPG keypair. A dialog box appears prompting you to create a new key pair. Enter your name, email address, and an optional comment. You can also choose an expiration time for your key, as well as the key strength (number of bits) and algorithms. The next dialog box prompts you for your passphrase. At this point, your key appears in the main KGpg window.

Warning If you forget your passphrase, the key cannot be used and any data encrypted using that key will be lost. To find your GPG key ID, look in the ''Key ID'' column next to the newly created key. In most cases, if you are asked for the key ID, you should prepend "0x" to the key ID, as in "0x6789ABCD". You should make a backup of your private key and store it somewhere secure.

3.9.3. Generating GPG Keys Using the Command Line Use the following shell command: gpg --gen-key This command generates a key pair that consists of a public and a private key. Other people use your public key to authenticate and/or decrypt your communications. Distribute your public key as widely as possible, especially to people who you know will want to receive authentic communications from you, such as a mailing list. The Fedora Documentation Project, for example, asks participants to include a GPG public key in their self-introduction.

144

Generating GPG Keys Using the Command Line

A series of prompts directs you through the process. Press the Enter key to assign a default value if desired. The first prompt asks you to select what kind of key you prefer: Please select what kind of key you want: (1) DSA and ElGamal (default) (2) DSA (sign only) (4) RSA (sign only) Your selection? In almost all cases, the default is the correct choice. A DSA/ElGamal key allows you not only to sign communications, but also to encrypt files. Next, choose the key size: minimum keysize is 768 bits default keysize is 1024 bits highest suggested keysize is 2048 bits What keysize do you want? (1024) Again, the default is sufficient for almost all users, and represents an ''extremely'' strong level of security. Next, choose when the key will expire. It is a good idea to choose an expiration date instead of using the default, which is ''none.'' If, for example, the email address on the key becomes invalid, an expiration date will remind others to stop using that public key. Please specify how long the key should be valid. 0 = key does not expire d = key expires in n days w = key expires in n weeks m = key expires in n months y = key expires in n years Key is valid for? (0) Entering a value of 1y, for example, makes the key valid for one year. (You may change this expiration date after the key is generated, if you change your mind.) Before the gpgcode> program asks for signature information, the following prompt appears: Is this correct (y/n)? Enter ycode> to finish the process. Next, enter your name and email address. Remember this process is about authenticating you as a real individual. For this reason, include your real name. Do not use aliases or handles, since these disguise or obfuscate your identity. Enter your real email address for your GPG key. If you choose a bogus email address, it will be more difficult for others to find your public key. This makes authenticating your communications difficult. If you are using this GPG key for [[DocsProject/SelfIntroduction| self-introduction]] on a mailing list, for example, enter the email address you use on that list. Use the comment field to include aliases or other information. (Some people use different keys for different purposes and identify each key with a comment, such as "Office" or "Open Source Projects.") At the confirmation prompt, enter the letter O to continue if all entries are correct, or use the other options to fix any problems. Finally, enter a passphrase for your secret key. The gpg program asks you to enter your passphrase twice to ensure you made no typing errors. Finally, gpg generates random data to make your key as unique as possible. Move your mouse, type random keys, or perform other tasks on the system during this step to speed up the process. Once this step is finished, your keys are complete and ready to use:

pub 1024D/1B2AFA1C 2005-03-31 John Q. Doe (Fedora Docs Project) Key fingerprint = 117C FE83 22EA B843 3E86 6486 4320 545E 1B2A FA1C sub 1024g/CEA4B22E 2005-03-31 [expires: 2006-03-31]

The key fingerprint is a shorthand "signature" for your key. It allows you to confirm to others that they have received your actual public key without any tampering. You do not need to write this fingerprint down. To display the fingerprint at any time, use this command, substituting your email address: gpg --fingerprint [email protected]

145


Your "GPG key ID" consists of 8 hex digits identifying the public key. In the example above, the GPG key ID is 1B2AFA1C. In most cases, if you are asked for the key ID, you should prepend "0x" to the key ID, as in "0x1B2AFA1C".

Warning If you forget your passphrase, the key cannot be used and any data encrypted using that key will be lost.

3.9.4. Using GPG with Alpine If you are using the email client Alpine or Pine then you will also need to download and install ez-pinegpg. This software is currently available from http://business-php.com/opensource/ez-pine-gpg/. Once you have installed ez-pine-gpg you will need to modify your ~/.pinerc file. You need to: 1. /home/username/bin should be replaced with the installation path that you specified. 2. In two places, the gpg-identifier after _RECIPIENTS_ should be replaced with your GPG public key's identifier. The reason you include your own GPG identifier here is so that if you send an encrypted message to "Alice", that message is also encrypted with your public key -- if you don't do this, then you will not be able to open that message in your sent-mail folder and remind yourself of what you wrote. It should look something like this:

# This variable takes a list of programs that message text is piped into # after MIME decoding, prior to display. display-filters=_LEADING("-----BEGIN PGP")_ /home/max/bin/ez-pine-gpg-incoming # This defines a program that message text is piped into before MIME # encoding, prior to sending sending-filters=/home/max/bin/ez-pine-gpg-sign _INCLUDEALLHDRS_, /home/username/bin/ez-pine-gpg-encrypt _RECIPIENTS_ gpg-identifier, /home/username/bin/ez-pine-gpg-sign-and-encrypt _INCLUDEALLHDRS_ _RECIPIENTS_ gpgidentifier

3.9.5. Using GPG with Evolution 3.9.5.1. Configuring GPG for use with Evolution To configure GPG for use in Evolution select from the Evolution Main Menu, select Tools, Settings... In the left pane, select Mail Accounts. In the right pane, select the email account you use for Fedora Project correspondence. Then select the Edit button. The Evolution Account Editor dialog appears. Select the Security tab. In the PGP/GPG Key ID field, enter the GPG key ID matching this account's email address. If you are not sure what your key ID is, use this command: gpg --fingerprint EMAIL_ADDRESS. The key ID is the same as the last eight characters (4 bytes) of the key fingerprint. It is a good idea to click the option Always encrypt to myself when sending encrypted mail. You may also want to select Always sign outgoing messages when using this account.

146

Using GPG with Thunderbird

Notice If you do not mark public keys as trusted in your keyring, you will not be able to encrypt email to their owners unless you select the option Always trust keys in my keyring when encrypting. You will instead receive a dialog indicating that a trust check has failed.

3.9.5.2. Verifying email with Evolution Evolution will automatically check any incoming GPG-signed messages for validity. If Evolution cannot GPG verify a message due to a missing public key (or tampering), it will end with a red banner. If the message is verified but you have not signed the key either locally or globally, the banner will be yellow. If the message is verified and you have signed the key, the banner will be green. When you click the seal icon, Evolution displays a dialog with more security information about the signature. To add a public key to your keyring, use the search function along with the key owner's email address: gpg -keyserver pgp.mit.edu --search email address. To import the correct key, you may need to match the key ID with the information provided by Evolution.

3.9.5.3. Signing and Encrypting email with Evolution Signing email allows the recipients to verify that the email actually came from you. The FDP (and the whole of the Fedora Project) encourage you to sign email to other participants, including on Fedora mailing lists. Encrypting email allows only your recipients to read your email. Please do not send encrypted email over the Fedora mailing lists, since almost no one will be able to read it. While composing your email, choose the Security menu, and then select PGP Sign to sign your message. To encrypt your message, select PGP Encrypt. You may sign an encrypted message as well, which is good practice. When you send the message, Evolution will ask you to enter your GPG key passphrase. (After three unsuccessful attempts Evolution generates an error.) If you select the option Remember this password for the remainder of this session, you will not need to use your passphrase again to sign or decrypt, unless you quit and restart Evolution.

3.9.6. Using GPG with Thunderbird Fedora Core includes Mozilla Thunderbird in the thunderbird package, and the mozilla-mail package for the Mozilla Suite email application. Thunderbird is the recommended Mozilla email application. This appears on your desktop as Applications > Internet > Thunderbird Email. Mozilla products support extensions, plugins that add new features to the main application. The Enigmail extensions provide GPG support to email products from Mozilla. Versions of Enigmail exist for both Mozilla Thunderbird, and the Mozilla Suite (Seamonkey). Netscape software from AOL is based on the Mozilla products, and may also use this extension. To install Enigmail on Fedora systems, follow the instructions given below. Enigmail uses the term OpenPGP in menu items and options. GPG is an implementation of OpenPGP, and you may treat the terms as equivalent. The homepage for Enigmail is: http://enigmail.mozdev.org/download.html. This page provides screenshots of Enigmail and GPG in action: http://enigmail.mozdev.org/ screenshots.html.

147


3.9.6.1. Installing Enigmail Enigmail is now available in fedora repository. It can be installed by typing: yum install thunderbird-enigmail at a command line. Alternatively, you can install thunderbird-enigmail using by going to System -> Administration -> Add/Remove Software.

3.9.7. About Public Key Encryption 1. Wikipedia - Public Key Cryptography 2. HowStuffWorks - Encryption

148

10

9

Chapter 4.

General Principles of Information Security The following general principals provide an overview of good security practices: • encrypt all data transmitted over networks to help prevent man-in-the-middle attacks and eavesdropping. It is important to encrypt authentication information, such as passwords. • minimize the amount of software installed and running services. • use security-enhancing software and tools, for example, Security-Enhanced Linux (SELinux) for Mandatory Access Control (MAC), Netfilter iptables for packet filtering (firewall), and the GNU Privacy Guard (GnuPG) for encrypting files. • if possible, run each network service on a separate system to minimize the risk of one compromised service being used to compromise other services. • maintain user accounts: create and enforce a strong password policy; delete unused user accounts. • routinely review system and application logs. By default, security-relevant system logs are written to /var/log/secure and /var/log/audit/audit.log. Note: sending logs to a dedicated log server helps prevent attackers from easily modifying local logs to avoid detection. • never log in as the root user unless absolutely necessary. It is recommended that administrators use sudo to execute commands as root when required. Users capable of running sudo are specified in /etc/sudoers. Use the visudo utility to edit /etc/sudoers.

4.1. Tips, Guides, and Tools 1

The United States' National Security Agency (NSA) provides hardening guides and tips for many different operating systems, to help government agencies, businesses, and individuals secure their systems against attack. The following guides (in PDF format) provide guidance for Red Hat Enterprise Linux 5: 2

• Hardening Tips for the Red Hat Enterprise Linux 5

• Guide to the Secure Configuration of Red Hat Enterprise Linux 5

3

4

The Defense Information Systems Agency (DISA) provides documentation, checklists, and tests 5 to help secure your system (Information Assurance Support Environment ). The UNIX SECURITY 6 TECHNICAL IMPLEMENTATION GUIDE (PDF) is a very specific guide to UNIX security - an advanced knowledge of UNIX and Linux is recommended before reading this guide. 7

The DISA UNIX Security Checklist Version 5, Release 1.16 provides a collection of documents and checklists, ranging from the correct ownerships and modes for system files, to patch control.

1

http://www.nsa.gov/ http://www.disa.mil/ 5 http://iase.disa.mil/index2.html 6 http://iase.disa.mil/stigs/stig/unix-stig-v5r1.pdf 7 http://iase.disa.mil/stigs/checklist/unix_checklist_v5r1-16_20090215.ZIP 4

149

Chapter 4. General Principles of Information Security

8

Also, DISA has made available UNIX SPR scripts that allow administrators to check specific settings on systems. These scripts provide XML-formatted reports listing any known vulnerable settings.

8

http://iase.disa.mil/stigs/SRR/unix.html

150

Chapter 5.

Secure Installation Security begins with the first time you put that CD or DVD into your disk drive to install Fedora. Configuring your system securely from the beginning makes it easier to implement additional security settings later.

5.1. Disk Partitions The NSA recommends creating separate partitions for /boot, /, /home, /tmp, and /var/tmp. The reasons for each are different and we will address each partition. /boot - This partition is the first partition that is read by the system during boot up. The boot loader and kernel images that are used to boot your system into Fedora are stored in this partition. This partition should not be encrypted. If this partition is included in / and that partition is encrypted or otherwise becomes unavailable then your system will not be able to boot. /home - When user data (/home) is stored in / instead of in a separate partition, the partition can fill up causing the operating system to become unstable. Also, when upgrading your system to the next version of Fedora it is a lot easier when you can keep your data in the /home partition as it will not be overwritten during installation. If the root partition (/) becomes corrupt your data could be lost forever. By using a separate partition there is slightly more protection against data loss. You can also target this partition for frequent backups. /tmp and /var/tmp - Both the /tmp and the /var/tmp directories are used to store data that doesn't need to be stored for a long period of time. However if a lot of data floods one of these directories it can consume all of your storage space. If this happens and these directories are stored within / then your system could become unstable and crash. For this reason, moving these directories into their own partitions is a good idea.

5.2. Utilize LUKS Partition Encryption 1

Since Fedora 9, implementation of Linux Unified Key Setup-on-disk-format (LUKS) encryption has become a lot easier. During the installation process an option to encrypt your partitions will be presented to the user. The user must supply a passphrase that will be the key to unlock the bulk encryption key that will be used to secure the partition's data.

1

http://fedoraproject.org/wiki/Security_Guide/9/LUKSDiskEncryption

151

152

Chapter 6.

Software Maintenance Software maintenance is extremely important to maintaining a secure system. It is vital to patch software as soon as it becomes available in order to prevent attackers from using known holes to infiltrate your system.

6.1. Install Minimal Software It is best practice to install only the packages you will use because each piece of software on your computer could possibly contain a vulnerability. If you are installing from the DVD media take the opportunity to select exactly what packages you want to install during the installation. When you find you need another package, you can always add it to the system later.

6.2. Plan and Configure Security Updates All software contains bugs. Often, these bugs can result in a vulnerability that can expose your system to malicious users. Unpatched systems are a common cause of computer intrusions. You should have a plan to install security patches in a timely manner to close those vulnerabilities so they can not be exploited. For home users, security updates should be installed as soon as possible. Configuring automatic installation of security updates is one way to avoid having to remember, but does carry a slight risk that something can cause a conflict with your configuration or with other software on the system. For business or advanced home users, security updates should be tested and schedule for installation. Additional controls will need to be used to protect the system during the time between the patch release and its installation on the system. These controls would depend on the exact vulnerability, but could include additional firewall rules, the use of external firewalls, or changes in software settings.

6.3. Adjusting Automatic Updates Fedora is configured to apply all updates on a daily schedule. If you want to change the how your system installs updates you must do so via '''Software Update Preferences'''. You can change the schedule, the type of updates to apply or to notify you of available updates. In Gnome, you can find controls for your updates at: System -> Preferences -> Software Updates. In KDE it is located at: Applications -> Settings -> Software Updates.

6.4. Install Signed Packages from Well Known Repositories Software packages are published through repositories. All well known repositories support package signing. Package signing uses public key technology to prove that the package that was published by the repository has not been changed since the signature was applied. This provides some protection against installing software that may have been maliciously altered after the package was created but before you downloaded it. Using too many repositories, untrustworthy repositories, or repositories with unsigned packages has a higher risk of introducing malicious or vulnerable code into your system. Use caution when adding repositories to yum/software update.

153

154

Chapter 7.

References The following references are pointers to additional information that is relevant to SELinux and Fedora but beyond the scope of this guide. Note that due to the rapid development of SELinux, some of this material may only apply to specific releases of Fedora. Books SELinux by Example Mayer, MacMillan, and Caplan Prentice Hall, 2007 Tutorials and Help Understanding and Customizing the Apache HTTP SELinux Policy http://fedora.redhat.com/docs/selinux-apache-fc3/ Tutorials and talks from Russell Coker http://www.coker.com.au/selinux/talks/ibmtu-2004/ Generic Writing SELinux policy HOWTO http://www.lurking-grue.org/writingselinuxpolicyHOWTO.html Red Hat Knowledgebase http://kbase.redhat.com/ General Information NSA SELinux main website 1 http://www.nsa.gov/selinux/ NSA SELinux FAQ 2 http://www.nsa.gov/selinux/info/faq.cfm Fedora SELinux FAQ http://fedora.redhat.com/docs/selinux-faq-fc3/ SELinux NSA's Open Source Security Enhanced Linux http://www.oreilly.com/catalog/selinux/ Technology An Overview of Object Classes and Permissions http://www.tresys.com/selinux/obj_perms_help.html Integrating Flexible Support for Security Policies into the Linux Operating System (a history of Flask implementation in Linux) http://www.nsa.gov/research/_files/selinux/papers/selsymp2005.pdf

1 2

http://www.nsa.gov/research/selinux/index.shtml http://www.nsa.gov/research/selinux/faqs.shtml

155

Chapter 7. References

Implementing SELinux as a Linux Security Module http://www.nsa.gov/research/_files/publications/implementing_selinux.pdf A Security Policy Configuration for the Security-Enhanced Linux http://www.nsa.gov/research/_files/selinux/papers/policy/policy.shtml Community Fedora SELinux User Guide http://docs.fedoraproject.org/selinux-user-guide/ Fedora SELinux Managing Confined Services Guide http://docs.fedoraproject.org/selinux-managing-confined-services-guide/ SELinux community page http://selinux.sourceforge.net IRC irc.freenode.net, #selinux, #fedora-selinux, #security History Quick history of Flask http://www.cs.utah.edu/flux/fluke/html/flask.html Full background on Fluke http://www.cs.utah.edu/flux/fluke/html/index.html

156

Appendix A. Encryption Standards A.1. Synchronous Encryption A.1.1. Advanced Encryption Standard - AES In cryptography, the Advanced Encryption Standard (AES) is an encryption standard adopted by the U.S. government. The standard comprises three block ciphers, AES-128, AES-192 and AES-256, adopted from a larger collection originally published as Rijndael. Each AES cipher has a 128-bit block size, with key sizes of 128, 192 and 256 bits, respectively. The AES ciphers have been analyzed extensively and are now used worldwide, as was the case with its predecessor, the Data Encryption 1 Standard (DES).

A.1.1.1. AES Uses A.1.1.2. AES History AES was announced by National Institute of Standards and Technology (NIST) as U.S. FIPS PUB 197 (FIPS 197) on November 26, 2001 after a 5-year standardization process in which fifteen competing designs were presented and evaluated before Rijndael was selected as the most suitable (see Advanced Encryption Standard process for more details). It became effective as a standard May 26, 2002. It is available in many different encryption packages. AES is the first publicly accessible and 2 open cipher approved by the NSA for top secret information (see Security of AES, below). The Rijndael cipher was developed by two Belgian cryptographers, Joan Daemen and Vincent Rijmen, and submitted by them to the AES selection process. Rijndael (pronounced [r�inda�l]) is a portmanteau 3 of the names of the two inventors.

A.1.2. Data Encryption Standard - DES The Data Encryption Standard (DES) is a block cipher (a form of shared secret encryption) that was selected by the National Bureau of Standards as an official Federal Information Processing Standard (FIPS) for the United States in 1976 and which has subsequently enjoyed widespread use internationally. It is based on a symmetric-key algorithm that uses a 56-bit key. The algorithm was initially controversial with classified design elements, a relatively short key length, and suspicions about a National Security Agency (NSA) backdoor. DES consequently came under intense academic 4 scrutiny which motivated the modern understanding of block ciphers and their cryptanalysis.

A.1.2.1. DES Uses A.1.2.2. DES History DES is now considered to be insecure for many applications. This is chiefly due to the 56-bit key size being too small; in January, 1999, distributed.net and the Electronic Frontier Foundation "Advanced Encryption Standard." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Advanced_Encryption_Standard "Advanced Encryption Standard." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Advanced_Encryption_Standard "Advanced Encryption Standard." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Advanced_Encryption_Standard "Data Encryption Standard." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Data_Encryption_Standard

157

Appendix A. Encryption Standards

collaborated to publicly break a DES key in 22 hours and 15 minutes (see chronology). There are also some analytical results which demonstrate theoretical weaknesses in the cipher, although they are unfeasible to mount in practice. The algorithm is believed to be practically secure in the form of Triple DES, although there are theoretical attacks. In recent years, the cipher has been superseded by the 5 Advanced Encryption Standard (AES). In some documentation, a distinction is made between DES as a standard and DES the algorithm which is referred to as the DEA (the Data Encryption Algorithm). When spoken, "DES" is either spelled 6 out as an abbreviation (/�di��i��s/), or pronounced as a one-syllable acronym (/�d�z/).

A.2. Public-key Encryption Public-key cryptography is a cryptographic approach, employed by many cryptographic algorithms and cryptosystems, whose distinguishing characteristic is the use of asymmetric key algorithms instead of or in addition to symmetric key algorithms. Using the techniques of public key-private key cryptography, many methods of protecting communications or authenticating messages formerly unknown have become practical. They do not require a secure initial exchange of one or more secret keys as is required when using symmetric key algorithms. It can also be used to create digital 7 signatures. Public key cryptography is a fundamental and widely used technology around the world, and is the approach which underlies such Internet standards as Transport Layer Security (TLS) (successor to 8 SSL), PGP and GPG. The distinguishing technique used in public key cryptography is the use of asymmetric key algorithms, where the key used to encrypt a message is not the same as the key used to decrypt it. Each user has a pair of cryptographic keys — a public key and a private key. The private key is kept secret, whilst the public key may be widely distributed. Messages are encrypted with the recipient's public key and can only be decrypted with the corresponding private key. The keys are related mathematically, but the private key cannot be feasibly (ie, in actual or projected practice) derived from the public key. It was the discovery of such algorithms which revolutionized the practice of cryptography beginning in the 9 middle 1970s. In contrast, Symmetric-key algorithms, variations of which have been used for some thousands of years, use a single secret key shared by sender and receiver (which must also be kept private, thus accounting for the ambiguity of the common terminology) for both encryption and decryption. To use a 10 symmetric encryption scheme, the sender and receiver must securely share a key in advance. Because symmetric key algorithms are nearly always much less computationally intensive, it is common to exchange a key using a key-exchange algorithm and transmit data using that key and a symmetric key algorithm. PGP, and the SSL/TLS family of schemes do this, for instance, and are 11 called hybrid cryptosystems in consequence.

A.2.1. Diffie-Hellman Diffie–Hellman key exchange (D–H) is a cryptographic protocol that allows two parties that have no prior knowledge of each other to jointly establish a shared secret key over an insecure "Data Encryption Standard." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Data_Encryption_Standard "Data Encryption Standard." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Data_Encryption_Standard "Public-key Encryption." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Public-key_cryptography "Public-key Encryption." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Public-key_cryptography "Public-key Encryption." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Public-key_cryptography "Public-key Encryption." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Public-key_cryptography "Public-key Encryption." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Public-key_cryptography

158

RSA

communications channel. This key can then be used to encrypt subsequent communications using a 12 symmetric key cipher.

A.2.1.1. Diffie-Hellman History The scheme was first published by Whitfield Diffie and Martin Hellman in 1976, although it later emerged that it had been separately invented a few years earlier within GCHQ, the British signals intelligence agency, by Malcolm J. Williamson but was kept classified. In 2002, Hellman suggested the algorithm be called Diffie–Hellman–Merkle key exchange in recognition of Ralph Merkle's contribution 13 to the invention of public-key cryptography (Hellman, 2002). Although Diffie–Hellman key agreement itself is an anonymous (non-authenticated) key-agreement protocol, it provides the basis for a variety of authenticated protocols, and is used to provide perfect forward secrecy in Transport Layer Security's ephemeral modes (referred to as EDH or DHE 14 depending on the cipher suite). U.S. Patent 4,200,770, now expired, describes the algorithm and credits Hellman, Diffie, and Merkle 15 as inventors.

A.2.2. RSA In cryptography, RSA (which stands for Rivest, Shamir and Adleman who first publicly described it; see below) is an algorithm for public-key cryptography. It is the first algorithm known to be suitable for signing as well as encryption, and was one of the first great advances in public key cryptography. RSA is widely used in electronic commerce protocols, and is believed to be secure given sufficiently long 16 keys and the use of up-to-date implementations.

A.2.3. DSA The Digital Signature Algorithm (DSA) is a United States Federal Government standard or FIPS for digital signatures. It was proposed by the National Institute of Standards and Technology (NIST) in August 1991 for use in their Digital Signature Standard (DSS), specified in FIPS 186, adopted in 1993. A minor revision was issued in 1996 as FIPS 186-1. The standard was expanded further in 2000 as 17 FIPS 186-2 and again in 2009 as FIPS 186-3.

A.2.4. SSL/TLS Transport Layer Security (TLS) and its predecessor, Secure Socket Layer (SSL), are cryptographic protocols that provide security for communications over networks such as the Internet. TLS and SSL encrypt the segments of network connections at the Transport Layer end-to-end. Several versions of the protocols are in widespread use in applications like web browsing, electronic mail, Internet faxing, instant messaging and voice-over-IP (VoIP). TLS is an IETF standards track protocol, last updated in RFC 5246, that was based on the earlier SSL specifications developed by Netscape Corporation. The TLS protocol allows client/server applications to communicate across a network in a way designed to prevent eavesdropping and tampering. TLS provides endpoint authentication and "Diffie-Hellman." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Diffie-Hellman "Diffie-Hellman." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Diffie-Hellman "Diffie-Hellman." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Diffie-Hellman "Diffie-Hellman." Wikipedia. 14 November 2009 http://en.wikipedia.org/wiki/Diffie-Hellman "RSA" Wikipedia 14 April 2010 http://en.wikipedia.org/wiki/RSA "Digital Signature Algorithm" Wikipedia 14 April 2010 http://en.wikipedia.org/wiki/Digital_Signature_Algorithm

159

Appendix A. Encryption Standards

communications confidentiality over the Internet using cryptography. TLS provides RSA security with 1024 and 2048 bit strengths. In typical end-user/browser usage, TLS authentication is unilateral: only the server is authenticated (the client knows the server's identity), but not vice versa (the client remains unauthenticated or anonymous). TLS also supports the more secure bilateral connection mode (typically used in enterprise applications), in which both ends of the "conversation" can be assured with whom they are communicating (provided they diligently scrutinize the identity information in the other party's certificate). This is known as mutual authentication, or 2SSL. Mutual authentication requires that the TLS client-side also hold a certificate (which is not usually the case in the end-user/browser scenario). Unless, that is, TLS-PSK, the Secure Remote Password (SRP) protocol, or some other protocol is used that can provide strong mutual authentication in the absence of certificates. Typically, the key information and certificates necessary for TLS are handled in the form of X.509 certificates, which define required fields and data formats. SSL operates in modular fashion. It is extensible by design, with support for forward and backward 18 compatibility and negotiation between peers.

A.2.5. Cramer-Shoup Cryptosystem The Cramer–Shoup system is an asymmetric key encryption algorithm, and was the first efficient scheme proven to be secure against adaptive chosen ciphertext attack using standard cryptographic assumptions. Its security is based on the computational intractability (widely assumed, but not proved) of the decisional Diffie–Hellman assumption. Developed by Ronald Cramer and Victor Shoup in 1998, it is an extension of the Elgamal cryptosystem. In contrast to Elgamal, which is extremely malleable, Cramer–Shoup adds additional elements to ensure non-malleability even against a resourceful attacker. This non-malleability is achieved through the use of a collision-resistant hash function and 19 additional computations, resulting in a ciphertext which is twice as large as in Elgamal.

A.2.6. ElGamal Encryption In cryptography, the ElGamal encryption system is an asymmetric key encryption algorithm for publickey cryptography which is based on the Diffie-Hellman key agreement. It was described by Taher Elgamal in 1985.[1] ElGamal encryption is used in the free GNU Privacy Guard software, recent versions of PGP, and other cryptosystems. The Digital Signature Algorithm is a variant of the ElGamal 20 signature scheme, which should not be confused with ElGamal encryption.

"Transport Layer Security" Wikipedia 14 April 2010 http://en.wikipedia.org/wiki/Transport_Layer_Security "Cramer–Shoup cryptosystem" Wikipedia 14 April 2010 http://en.wikipedia.org/wiki/Cramer-Shoup_cryptosystem "ElGamal encryption" Wikipedia 14 April 2010 http://en.wikipedia.org/wiki/ElGamal_encryption

160

Appendix B. Revision History Revision Wed Jun 9 2010 Eric Christensen 13.0-8 [email protected] Removed FireGPG text as that project has been retired.

Revision Fri May 14 2010 Eric Christensen 13.0-7 [email protected] Removed "bug" text from 7-Zip chapter per bug 591980.

Revision Wed Apr 14 2010 Eric Christensen 13.0-6 [email protected] Completed the encryption standards appendix.

Revision Fri Apr 09 2010 13.0-5 Added "Using GPG with Alpine". Added "Using GPG with Evolution".

Eric Christensen [email protected]

Revision Tue Apr 06 2010 Eric Christensen 13.0-4 [email protected] Repaired issues regarding untranslatable text in para.

Revision Tue Apr 06 2010 Eric Christensen 13.0-3 [email protected] Removed the PackageKit vulnerability text seen in Fedora 12.

Revision Fri Nov 20 2009 Eric Christensen 13.0-2 [email protected] Added the Revision History to the end of the document. Added the Encryption Standards appendix.

Revision Fri Nov 20 2009 13.0-1 Fedora 13 branch.


Revision Thu Nov 19 2009 Eric Christensen 1.0-23 [email protected] Updated the section "Local users may install trusted packages" to the latest fix, again.

161

Appendix B. Revision History

Revision Thu Nov 19 2009 Eric Christensen 1.0-22 [email protected] Updated the section "Local users may install trusted packages" to the latest fix.

Revision Wed Nov 18 2009 Eric Christensen 1.0-21 [email protected] Added section "Local users may install trusted packages".

Revision Sat Nov 14 2009 Eric Christensen 1.0-20 [email protected] Added information from Wikipedia to the Encryption Standards appendix. Added Adam Ligas to the author page for his role in developing the 7-Zip portions.

Revision Mon Oct 26 2009 1.0-19 Updated license to CC-BY-SA.


Revision Wed Aug 05 2009 1.0-18 Fixed issues related to Bug 515043.

Eric Chrisetnsen [email protected]

Revision Mon Jul 27 2009 1.0-17 Repaired vendor information in SPEC.


Revision Fri Jul 24 2009 Fedora Release Engineering 1.0-16 [email protected] Rebuilt for https://fedoraproject.org/wiki/Fedora_12_Mass_Rebuild

Revision Tue Jul 14 2009 Eric Christensen 1.0-15 [email protected] Added "desktop-file-utils" to BUILDREQUIRES on the spec

Revision Tue Mar 10 2009 Scott Radvan [email protected] 1.0-14 Remove more rhel specifics, major review and remove draft, ready for push

Revision 1.0-13

162

Mon Mar 2 2009

Scott Radvan [email protected]

Lots of minor fixes

Revision 1.0-12

Wed Feb 11 2009

Scott Radvan [email protected]

new screenshots from F11 replacing existing/older ones

Revision Tue Feb 03 2009 Scott Radvan [email protected] 1.0-11 LUKS specifics to Fedora 9 modified to include later releases as well. Fix 404s in reference section, mainly bad NSA links. minor formatting changes.

Revision Wed Jan 27 2009 1.0-10 Fixed missing firewall setup screenshot.


Revision 1.0-9 Wed Jan 27 2009

Eric Christensen [email protected] Repaired items found to be incorrect during validation. Many Red Hat references have been changed to Fedora references.

163

164

Security Guide - A Guide to Securing Fedora Linux

Security Guide - A Guide to Securing Fedora Linux

Suggest Documents